fix: harden bluebubbles short ids and fetch wrapper (#1369 ) (thanks @tyler6204)

feat: enhance message context with full ID support for replies and caching
- Updated message processing to include full message IDs alongside short IDs for better context resolution. - Improved reply handling by caching inbound messages, allowing for accurate sender and body resolution without exposing dropped content. - Adjusted tests to validate the new full ID properties and their integration into the message handling workflow.
2026-01-21 17:05:36 +00:00 · 2026-01-21 16:50:29 +00:00 · 2026-01-21 16:50:28 +00:00 · 2026-01-21 16:50:28 +00:00 · 2026-01-21 16:48:42 +00:00 · 2026-01-21 16:48:42 +00:00
123 changed files with 3827 additions and 1532 deletions
@@ -2,226 +2,202 @@

 Docs: https://docs.clawd.bot

-## 2026.1.19-3
+## 2026.1.21

 ### Changes
- Android: remove legacy bridge transport code now that nodes use the gateway protocol.
- Android: send structured payloads in node events/invokes and include user-agent metadata in gateway connects.
+- CLI: default exec approvals to the local host, add gateway/node targeting flags, and show target details in allowlist output.
+- CLI: exec approvals mutations render tables instead of raw JSON.
+- Exec approvals: support wildcard agent allowlists (`*`) across all agents.
+- Nodes: expose node PATH in status/describe and bootstrap PATH for node-host execution.
+- CLI: flatten node service commands under `clawdbot node` and remove `service node` docs.

 ### Fixes
- Slack: respect verbose tool summaries and keep tool notifications threaded. (#1360) — thanks @SocialNerd42069.
+- Nodes tool: include agent/node/gateway context in tool failure logs to speed approval debugging.
+- macOS: exec approvals now respect wildcard agent allowlists (`*`).
+- UI: remove the chat stop button and keep the composer aligned to the bottom edge.
+- Configure: restrict the model allowlist picker to OAuth-compatible Anthropic models and preselect Opus 4.5.
+- Configure: seed model fallbacks from the allowlist selection when multiple models are chosen.
+- Model picker: list the full catalog when no model allowlist is configured.
+- BlueBubbles: resolve short message IDs safely and expose full IDs in templates. (#1369) Thanks @tyler6204.
+- Infra: preserve fetch helper methods when wrapping abort signals. (#1369)

-## 2026.1.19-2
+## 2026.1.20

 ### Changes
- Android: migrate node transport to the Gateway WebSocket protocol with TLS pinning support + gateway discovery naming.
- Android: bump okhttp + dnsjava to satisfy lint dependency checks.
- Docs: refresh Android node discovery docs for the Gateway WS service type.
-
-### Fixes
- Tests: stabilize Windows gateway/CLI tests by skipping sidecars, normalizing argv, and extending timeouts.
- CLI: skip runner rebuilds when dist is fresh. (#1231) — thanks @mukhtharcm, @thewilloftheshadow.
-
-## 2026.1.19-1
-
-### Breaking
- **BREAKING:** Reject invalid/unknown config entries and refuse to start the gateway for safety; run `clawdbot doctor --fix` to repair. 
-
-### Changes
- Usage: add `/usage cost` summaries and macOS menu cost submenu with daily charting.
- Agents: clarify node_modules read-only guidance in agent instructions.
- TUI: add syntax highlighting for code blocks. (#1200) — thanks @vignesh07.
-
-### Fixes
- UI: enable shell mode for sync Windows spawns to avoid `pnpm ui:build` EINVAL. (#1212) — thanks @longmaba.
- Agents: add `clawdbot agents set-identity` helper and update bootstrap guidance for multi-agent setups. (#1222) — thanks @ThePickle31.
- Plugins: surface plugin load/register/config errors in gateway logs with plugin/source context.
- Agents: propagate accountId into embedded runs so sub-agent announce routing honors the originating account. (#1058)
- Compaction: include tool failure summaries in safeguard compaction to prevent retry loops. (#1084)
- Daemon: include HOME in service environments to avoid missing HOME errors. (#1214) — thanks @ameno-.
- TUI: show generic empty-state text for searchable pickers. (#1201) — thanks @vignesh07.
- Doctor: canonicalize legacy session keys in session stores to prevent stale metadata. (#1169)
- CLI: centralize CLI command registration to keep fast-path routing and program wiring in sync. (#1207) — thanks @gumadeiras.
-
-## 2026.1.18-5
-
-### Changes
- Dependencies: update core + plugin deps (grammy, vitest, openai, Microsoft agents hosting, etc.).
- Onboarding: add allowlist prompts and username-to-id resolution across core and extension channels.
- TUI: add searchable model picker for quicker model selection. (#1198) — thanks @vignesh07.
- Docs: clarify allowlist input types and onboarding behavior for messaging channels.
-
-### Fixes
- Configure: hide OpenRouter auto routing model from the model picker. (#1182) — thanks @zerone0x.
- Docs: make docs:list fail fast with a clear error if the docs directory is missing.
- macOS: load menu session previews asynchronously so items populate while the menu is open.
- macOS: use label colors for session preview text so previews render in menu subviews.
- macOS: suppress usage error text in the menubar cost view.
- Telegram: honor pairing allowlists for native slash commands.
- TUI: highlight model search matches and stabilize search ordering.
- CLI: keep banners on routed commands, restore config guarding outside fast-path routing, and tighten fast-path flag parsing while skipping console capture for extra speed. (#1195) — thanks @gumadeiras.
- Slack: resolve Bolt import interop for Bun + Node. (#1191) — thanks @CoreyH.
- Gateway: require authorized restarts for SIGUSR1 (restart/apply/update) so config gating can't be bypassed.
- Discord: stop reconnecting the gateway after aborts to prevent duplicate listeners.
-
-## 2026.1.18-4
-
-### Changes
- macOS: switch PeekabooBridge integration to the tagged Swift Package Manager release (no submodule).
- macOS: stop syncing Peekaboo as a git submodule in postinstall.
- Swabble: use the tagged Commander Swift package release.
- CLI: add `clawdbot acp client` interactive ACP harness for debugging.
- Plugins: route command detection/text chunking helpers through the plugin runtime and drop runtime exports from the SDK.
+- Control UI: add copy-as-markdown with error feedback. (#1345) https://docs.clawd.bot/web/control-ui
+- Control UI: drop the legacy list view. (#1345) https://docs.clawd.bot/web/control-ui
+- TUI: add syntax highlighting for code blocks. (#1200) https://docs.clawd.bot/tui
+- TUI: session picker shows derived titles, fuzzy search, relative times, and last message preview. (#1271) https://docs.clawd.bot/tui
+- TUI: add a searchable model picker for quicker model selection. (#1198) https://docs.clawd.bot/tui
+- TUI: add input history (up/down) for submitted messages. (#1348) https://docs.clawd.bot/tui
+- ACP: add `clawdbot acp` for IDE integrations. https://docs.clawd.bot/cli/acp
+- ACP: add `clawdbot acp client` interactive harness for debugging. https://docs.clawd.bot/cli/acp
+- Skills: add download installs with OS-filtered options. https://docs.clawd.bot/tools/skills
+- Skills: add the local sherpa-onnx-tts skill. https://docs.clawd.bot/tools/skills
+- Memory: add hybrid BM25 + vector search (FTS5) with weighted merging and fallback. https://docs.clawd.bot/concepts/memory
+- Memory: add SQLite embedding cache to speed up reindexing and frequent updates. https://docs.clawd.bot/concepts/memory
+- Memory: add OpenAI batch indexing for embeddings when configured. https://docs.clawd.bot/concepts/memory
+- Memory: enable OpenAI batch indexing by default for OpenAI embeddings. https://docs.clawd.bot/concepts/memory
+- Memory: allow parallel OpenAI batch indexing jobs (default concurrency: 2). https://docs.clawd.bot/concepts/memory
+- Memory: render progress immediately, color batch statuses in verbose logs, and poll OpenAI batch status every 2s by default. https://docs.clawd.bot/concepts/memory
+- Memory: add `--verbose` logging for memory status + batch indexing details. https://docs.clawd.bot/concepts/memory
+- Memory: add native Gemini embeddings provider for memory search. (#1151) https://docs.clawd.bot/concepts/memory
+- Browser: allow config defaults for efficient snapshots in the tool/CLI. (#1336) https://docs.clawd.bot/tools/browser
+- Nostr: add the Nostr channel plugin with profile management + onboarding defaults. (#1323) https://docs.clawd.bot/channels/nostr
+- Matrix: migrate to matrix-bot-sdk with E2EE support, location handling, and group allowlist upgrades. (#1298) https://docs.clawd.bot/channels/matrix
+- Slack: add HTTP webhook mode via Bolt HTTP receiver. (#1143) https://docs.clawd.bot/channels/slack
+- Telegram: enrich forwarded-message context with normalized origin details + legacy fallback. (#1090) https://docs.clawd.bot/channels/telegram
+- Discord: fall back to `/skill` when native command limits are exceeded. (#1287)
+- Discord: expose `/skill` globally. (#1287)
+- Zalouser: add channel dock metadata, config schema, setup wiring, probe, and status issues. (#1219) https://docs.clawd.bot/plugins/zalouser
+- Plugins: require manifest-embedded config schemas with preflight validation warnings. (#1272) https://docs.clawd.bot/plugins/manifest
+- Plugins: move channel catalog metadata into plugin manifests. (#1290) https://docs.clawd.bot/plugins/manifest
+- Plugins: align Nextcloud Talk policy helpers with core patterns. (#1290) https://docs.clawd.bot/plugins/manifest
+- Plugins/UI: let channel plugin metadata drive UI labels/icons and cron channel options. (#1306) https://docs.clawd.bot/web/control-ui
+- Plugins: add plugin slots with a dedicated memory slot selector. https://docs.clawd.bot/plugins/agent-tools
+- Plugins: ship the bundled BlueBubbles channel plugin (disabled by default). https://docs.clawd.bot/channels/bluebubbles
+- Plugins: migrate bundled messaging extensions to the plugin SDK and resolve plugin-sdk imports in the loader.
+- Plugins: migrate the Zalo plugin to the shared plugin SDK runtime. https://docs.clawd.bot/channels/zalo
+- Plugins: migrate the Zalo Personal plugin to the shared plugin SDK runtime. https://docs.clawd.bot/plugins/zalouser
+- Plugins: allow optional agent tools with explicit allowlists and add the plugin tool authoring guide. https://docs.clawd.bot/plugins/agent-tools
 - Plugins: auto-enable bundled channel/provider plugins when configuration is present.
+- Plugins: sync plugin sources on channel switches and update npm-installed plugins during `clawdbot update`.
+- Plugins: share npm plugin update logic between `clawdbot update` and `clawdbot plugins update`.
+- Gateway/API: add `/v1/responses` (OpenResponses) with item-based input + semantic streaming events. (#1229)
+- Gateway/API: expand `/v1/responses` to support file/image inputs, tool_choice, usage, and output limits. (#1229)
+- Usage: add `/usage cost` summaries and macOS menu cost charts. https://docs.clawd.bot/reference/api-usage-costs
+- Security: warn when <=300B models run without sandboxing while web tools are enabled. https://docs.clawd.bot/cli/security
+- Exec: add host/security/ask routing for gateway + node exec. https://docs.clawd.bot/tools/exec
+- Exec: add `/exec` directive for per-session exec defaults (host/security/ask/node). https://docs.clawd.bot/tools/exec
+- Exec approvals: migrate approvals to `~/.clawdbot/exec-approvals.json` with per-agent allowlists + skill auto-allow toggle, and add approvals UI + node exec lifecycle events. https://docs.clawd.bot/tools/exec-approvals
+- Nodes: add headless node host (`clawdbot node start`) for `system.run`/`system.which`. https://docs.clawd.bot/cli/node
+- Nodes: add node daemon service install/status/start/stop/restart. https://docs.clawd.bot/cli/node
+- Bridge: add `skills.bins` RPC to support node host auto-allow skill bins.
+- Sessions: add daily reset policy with per-type overrides and idle windows (default 4am local), preserving legacy idle-only configs. (#1146) https://docs.clawd.bot/concepts/session
+- Sessions: allow `sessions_spawn` to override thinking level for sub-agent runs. https://docs.clawd.bot/tools/subagents
+- Channels: unify thread/topic allowlist matching + command/mention gating helpers across core providers. https://docs.clawd.bot/concepts/groups
+- Models: add Qwen Portal OAuth provider support. (#1120) https://docs.clawd.bot/providers/qwen
+- Onboarding: add allowlist prompts and username-to-id resolution across core and extension channels. https://docs.clawd.bot/start/onboarding
+- Docs: clarify allowlist input types and onboarding behavior for messaging channels. https://docs.clawd.bot/start/onboarding
+- Docs: refresh Android node discovery docs for the Gateway WS service type. https://docs.clawd.bot/platforms/android
+- Docs: surface Amazon Bedrock in provider lists and clarify Bedrock auth env vars. (#1289) https://docs.clawd.bot/bedrock
+- Docs: clarify WhatsApp voice notes. https://docs.clawd.bot/channels/whatsapp
+- Docs: clarify Windows WSL portproxy LAN access notes. https://docs.clawd.bot/platforms/windows
+- Docs: refresh bird skill install metadata and usage notes. (#1302) https://docs.clawd.bot/tools/browser-login
+- Agents: add local docs path resolution and include docs/mirror/source/community pointers in the system prompt.
+- Agents: clarify node_modules read-only guidance in agent instructions.
 - Config: stamp last-touched metadata on write and warn if the config is newer than the running build.
 - macOS: hide usage section when usage is unavailable instead of showing provider errors.
- Memory: add native Gemini embeddings provider for memory search. (#1151)
- Agents: add local docs path resolution and include docs/mirror/source/community pointers in the system prompt.
- Slack: add HTTP webhook mode via Bolt HTTP receiver for Events API deployments. (#1143) — thanks @jdrhyne.
+- Android: migrate node transport to the Gateway WebSocket protocol with TLS pinning support + gateway discovery naming.
+- Android: send structured payloads in node events/invokes and include user-agent metadata in gateway connects.
+- Android: remove legacy bridge transport code now that nodes use the gateway protocol.
+- Android: bump okhttp + dnsjava to satisfy lint dependency checks.
+- Build: update workspace + core/plugin deps.
+- Build: use tsgo for dev/watch builds by default (opt out with `CLAWDBOT_TS_COMPILER=tsc`).
+- Repo: remove the Peekaboo git submodule now that the SPM release is used.
+- macOS: switch PeekabooBridge integration to the tagged Swift Package Manager release.
+- macOS: stop syncing Peekaboo in postinstall.
+- Swabble: use the tagged Commander Swift package release.
+
+### Breaking
+- **BREAKING:** Reject invalid/unknown config entries and refuse to start the gateway for safety. Run `clawdbot doctor --fix` to repair, then update plugins (`clawdbot plugins update`) if you use any.

 ### Fixes
- Auth profiles: keep auto-pinned preference while allowing rotation on failover; user pins stay locked. (#1138) — thanks @cheeeee.
+- Discovery: shorten Bonjour DNS-SD service type to `_clawdbot-gw._tcp` and update discovery clients/docs.
+- Diagnostics: export OTLP logs, correct queue depth tracking, and document message-flow telemetry.
+- Diagnostics: emit message-flow diagnostics across channels via shared dispatch. (#1244)
+- Diagnostics: gate heartbeat/webhook logging. (#1244)
+- Gateway: strip inbound envelope headers from chat history messages to keep clients clean.
+- Gateway: clarify unauthorized handshake responses with token/password mismatch guidance.
+- Gateway: allow mobile node client ids for iOS + Android handshake validation. (#1354)
+- Gateway: clarify connect/validation errors for gateway params. (#1347)
+- Gateway: preserve restart wake routing + thread replies across restarts. (#1337)
+- Gateway: reschedule per-agent heartbeats on config hot reload without restarting the runner.
+- Gateway: require authorized restarts for SIGUSR1 (restart/apply/update) so config gating can't be bypassed.
+- Cron: auto-deliver isolated agent output to explicit targets without tool calls. (#1285)
+- Agents: preserve subagent announce thread/topic routing + queued replies across channels. (#1241)
+- Agents: propagate accountId into embedded runs so sub-agent announce routing honors the originating account. (#1058)
+- Agents: avoid treating timeout errors with "aborted" messages as user aborts, so model fallback still runs. (#1137)
 - Agents: sanitize oversized image payloads before send and surface image-dimension errors.
- macOS: Doctor repairs LaunchAgent bootstrap issues for Gateway + Node when listed but not loaded. (#1166) — thanks @AlexMikhalev.
- macOS: avoid touching launchd in Remote over SSH so quitting the app no longer disables the remote gateway. (#1105)
+- Sessions: fall back to session labels when listing display names. (#1124)
+- Compaction: include tool failure summaries in safeguard compaction to prevent retry loops. (#1084)
+- Config: log invalid config issues once per run and keep invalid-config errors stackless.
+- Config: allow Perplexity as a web_search provider in config validation. (#1230)
+- Config: allow custom fields under `skills.entries.<name>.config` for skill credentials/config. (#1226)
+- Doctor: clarify plugin auto-enable hint text in the startup banner.
+- Doctor: canonicalize legacy session keys in session stores to prevent stale metadata. (#1169)
+- Docs: make docs:list fail fast with a clear error if the docs directory is missing.
+- Plugins: add Nextcloud Talk manifest for plugin config validation. (#1297)
+- Plugins: surface plugin load/register/config errors in gateway logs with plugin/source context.
+- CLI: preserve cron delivery settings when editing message payloads. (#1322)
+- CLI: keep `clawdbot logs` output resilient to broken pipes while preserving progress output.
+- CLI: avoid duplicating --profile/--dev flags when formatting commands.
+- CLI: centralize CLI command registration to keep fast-path routing and program wiring in sync. (#1207)
+- CLI: keep banners on routed commands, restore config guarding outside fast-path routing, and tighten fast-path flag parsing while skipping console capture for extra speed. (#1195)
+- CLI: skip runner rebuilds when dist is fresh. (#1231)
+- CLI: add WSL2/systemd unavailable hints in daemon status/doctor output.
+- Status: route native `/status` to the active agent so model selection reflects the correct profile. (#1301)
+- Status: show both usage windows with reset hints when usage data is available. (#1101)
+- UI: keep config form enums typed, preserve empty strings, protect sensitive defaults, and deepen config search. (#1315)
+- UI: preserve ordered list numbering in chat markdown. (#1341)
+- UI: allow Control UI to read gatewayUrl from URL params for remote WebSocket targets. (#1342)
+- UI: prevent double-scroll in Control UI chat by locking chat layout to the viewport. (#1283)
+- UI: enable shell mode for sync Windows spawns to avoid `pnpm ui:build` EINVAL. (#1212)
+- TUI: keep thinking blocks ordered before content during streaming and isolate per-run assembly. (#1202)
+- TUI: align custom editor initialization with the latest pi-tui API. (#1298)
+- TUI: show generic empty-state text for searchable pickers. (#1201)
+- TUI: highlight model search matches and stabilize search ordering.
+- Configure: hide OpenRouter auto routing model from the model picker. (#1182)
+- Memory: show total file counts + scan issues in `clawdbot memory status`.
+- Memory: fall back to non-batch embeddings after repeated batch failures.
+- Memory: apply OpenAI batch defaults even without explicit remote config.
 - Memory: index atomically so failed reindex preserves the previous memory database. (#1151)
 - Memory: avoid sqlite-vec unique constraint failures when reindexing duplicate chunk ids. (#1151)
-
-## 2026.1.18-3
-
-### Changes
- Exec: add host/security/ask routing for gateway + node exec.
- Exec: add `/exec` directive for per-session exec defaults (host/security/ask/node).
- macOS: migrate exec approvals to `~/.clawdbot/exec-approvals.json` with per-agent allowlists and skill auto-allow toggle.
- macOS: add approvals socket UI server + node exec lifecycle events.
- Nodes: add headless node host (`clawdbot node start`) for `system.run`/`system.which`.
- Nodes: add node daemon service install/status/start/stop/restart.
- Bridge: add `skills.bins` RPC to support node host auto-allow skill bins.
- Slash commands: replace `/cost` with `/usage off|tokens|full` to control per-response usage footer; `/usage` no longer aliases `/status`. (Supersedes #1140) — thanks @Nachx639.
- Sessions: add daily reset policy with per-type overrides and idle windows (default 4am local), preserving legacy idle-only configs. (#1146) — thanks @austinm911.
- Agents: auto-inject local image references for vision models and avoid reloading history images. (#1098) — thanks @tyler6204.
- Docs: refresh exec/elevated/exec-approvals docs for the new flow. https://docs.clawd.bot/tools/exec-approvals
- Docs: add node host CLI + update exec approvals/bridge protocol docs. https://docs.clawd.bot/cli/node
- ACP: add experimental ACP support for IDE integrations (`clawdbot acp`). Thanks @visionik.
- Tools: allow `sessions_spawn` to override thinking level for sub-agent runs.
- Channels: unify thread/topic allowlist matching + command/mention gating helpers across core providers.
- Models: add Qwen Portal OAuth provider support. (#1120) — thanks @mukhtharcm.
- Memory: add `--verbose` logging for memory status + batch indexing details.
- Memory: allow parallel OpenAI batch indexing jobs (default concurrency: 2).
- macOS: add per-agent exec approvals with allowlists, skill CLI auto-allow, and settings UI.
- Docs: add exec approvals guide and link from tools index. https://docs.clawd.bot/tools/exec-approvals
- macOS: add exec-host IPC for node service `system.run` with HMAC + peer UID checks.
-
-### Fixes
- Exec approvals: enforce allowlist when ask is off; prefer raw command for node approvals/events.
- Tools: return a companion-app-required message when node exec is requested with no paired node.
- Streaming: emit assistant deltas for OpenAI-compatible SSE chunks. (#1147) — thanks @alauppe.
- Model fallback: treat timeout aborts as failover while preserving user aborts. (#1137) — thanks @cheeeee.
-
-## 2026.1.18-2
-
-### Fixes
- Tests: stabilize plugin SDK resolution and embedded agent timeouts.
-
-## 2026.1.18-1
-
-### Changes
- Tools: allow `sessions_spawn` to override thinking level for sub-agent runs.
- Channels: unify thread/topic allowlist matching + command/mention gating helpers across core providers.
- Models: add Qwen Portal OAuth provider support. (#1120) — thanks @mukhtharcm.
- Memory: add `--verbose` logging for memory status + batch indexing details.
- Memory: allow parallel OpenAI batch indexing jobs (default concurrency: 2).
- macOS: add per-agent exec approvals with allowlists, skill CLI auto-allow, and settings UI.
- Docs: add exec approvals guide and link from tools index. https://docs.clawd.bot/tools/exec-approvals
-
-### Fixes
- Memory: apply OpenAI batch defaults even without explicit remote config.
- macOS: bundle Textual resources in packaged app builds to avoid code block crashes. (#1006)
- Tools: return a companion-app-required message when `system.run` is requested without a supporting node.
- Discord: only emit slow listener warnings after 30s.
-
-## 2026.1.17-6
-
-### Changes
- Plugins: add exclusive plugin slots with a dedicated memory slot selector.
- Memory: ship core memory tools + CLI as the bundled `memory-core` plugin.
- Docs: document plugin slots and memory plugin behavior.
- Plugins: add the bundled BlueBubbles channel plugin (disabled by default).
- Plugins: migrate bundled messaging extensions to the plugin SDK; resolve plugin-sdk imports in loader.
- Plugins: migrate the Zalo plugin to the shared plugin SDK runtime.
- Plugins: migrate the Zalo Personal plugin to the shared plugin SDK runtime.
-
-## 2026.1.17-5
-
-### Changes
- Memory: add hybrid BM25 + vector search (FTS5) with weighted merging and fallback.
- Memory: add SQLite embedding cache to speed up reindexing and frequent updates.
- CLI: surface FTS + embedding cache state in `clawdbot memory status`.
- Memory: render progress immediately, color batch statuses in verbose logs, and poll OpenAI batch status every 2s by default.
- Plugins: allow optional agent tools with explicit allowlists and add plugin tool authoring guide. https://docs.clawd.bot/plugins/agent-tools
- Tools: centralize plugin tool policy helpers.
- Commands: add `/subagents info` and show sub-agent counts in `/status`.
- Docs: clarify plugin agent tool configuration. https://docs.clawd.bot/plugins/agent-tools
-
-### Fixes
- Voice call: include request query in Twilio webhook verification when publicUrl is set. (#864)
-
-## 2026.1.18-1
-
-### Changes
- Tools: allow `sessions_spawn` to override thinking level for sub-agent runs.
- Channels: unify thread/topic allowlist matching + command/mention gating helpers across core providers.
- Models: add Qwen Portal OAuth provider support. (#1120) — thanks @mukhtharcm.
- Memory: add `--verbose` logging for memory status + batch indexing details.
- Memory: allow parallel OpenAI batch indexing jobs (default concurrency: 2).
- macOS: add per-agent exec approvals with allowlists, skill CLI auto-allow, and settings UI.
- Docs: add exec approvals guide and link from tools index. https://docs.clawd.bot/tools/exec-approvals
-
-### Fixes
- Memory: apply OpenAI batch defaults even without explicit remote config.
- macOS: bundle Textual resources in packaged app builds to avoid code block crashes. (#1006)
- Tools: return a companion-app-required message when `system.run` is requested without a supporting node.
- Discord: only emit slow listener warnings after 30s.
-## 2026.1.17-3
-
-### Changes
- Memory: add OpenAI Batch API indexing for embeddings when configured.
- Memory: enable OpenAI batch indexing by default for OpenAI embeddings.
-
-### Fixes
 - Memory: retry transient 5xx errors (Cloudflare) during embedding indexing.
-
-## 2026.1.17-2
-
-### Changes
-
-### Fixes
- Tools: show exec elevated flag before the command and keep it outside markdown in tool summaries.
 - Memory: parallelize embedding indexing with rate-limit retries.
 - Memory: split overly long lines to keep embeddings under token limits.
 - Memory: skip empty chunks to avoid invalid embedding inputs.
- Sessions: fall back to session labels when listing display names. (#1124) — thanks @abdaraxus.
- Discord: inherit parent channel allowlists for thread slash commands and reactions. (#1123) — thanks @thewilloftheshadow.
-
-## 2026.1.17-1
-
-### Changes
- Telegram: enrich forwarded message context with normalized origin details + legacy fallback. (#1090) — thanks @sleontenko.
- macOS: strip prerelease/build suffixes when parsing gateway semver patches. (#1110) — thanks @zerone0x.
- macOS: keep CLI install pinned to the full build suffix. (#1111) — thanks @artuskg.
- CLI: surface update availability in `clawdbot status`.
- CLI: add `clawdbot memory status --deep/--index` probes.
- CLI: add playful update completion quips.
-
-### Fixes
- Doctor: avoid re-adding WhatsApp ack reaction config when only legacy auth files exist. (#1087) — thanks @YuriNachos.
- Hooks: parse multi-line/YAML frontmatter metadata blocks (JSON5-friendly). (#1114) — thanks @sebslight.
- CLI: add WSL2/systemd unavailable hints in daemon status/doctor output.
- Windows: install gateway scheduled task as the current user; show friendly guidance instead of failing on access denied.
- Status: show both usage windows with reset hints when usage data is available. (#1101) — thanks @rhjoh.
- Memory: probe sqlite-vec availability in `clawdbot memory status`.
 - Memory: split embedding batches to avoid OpenAI token limits during indexing.
- Telegram: preserve hidden text_link URLs by expanding entities in inbound text. (#1118) — thanks @sleontenko.
+- Memory: probe sqlite-vec availability in `clawdbot memory status`.
+- Exec approvals: enforce allowlist when ask is off.
+- Exec approvals: prefer raw command for node approvals/events.
+- Tools: show exec elevated flag before the command and keep it outside markdown in tool summaries.
+- Tools: return a companion-app-required message when node exec is requested with no paired node.
+- Tools: return a companion-app-required message when `system.run` is requested without a supporting node.
+- Exec: default gateway/node exec security to allowlist when unset (sandbox stays deny).
+- Exec: prefer bash when fish is default shell, falling back to sh if bash is missing. (#1297)
+- Exec: merge login-shell PATH for host=gateway exec while keeping daemon PATH minimal. (#1304)
+- Streaming: emit assistant deltas for OpenAI-compatible SSE chunks. (#1147)
+- Discord: make resolve warnings avoid raw JSON payloads on rate limits.
+- Discord: process message handlers in parallel across sessions to avoid event queue blocking. (#1295)
+- Discord: stop reconnecting the gateway after aborts to prevent duplicate listeners.
+- Discord: only emit slow listener warnings after 30s.
+- Discord: inherit parent channel allowlists for thread slash commands and reactions. (#1123)
+- Telegram: honor pairing allowlists for native slash commands.
+- Telegram: preserve hidden text_link URLs by expanding entities in inbound text. (#1118)
+- Slack: resolve Bolt import interop for Bun + Node. (#1191)
+- Web search: infer Perplexity base URL from API key source (direct vs OpenRouter).
+- Web fetch: harden SSRF protection with shared hostname checks and redirect limits. (#1346)
+- Browser: register AI snapshot refs for act commands. (#1282)
+- Voice call: include request query in Twilio webhook verification when publicUrl is set. (#864)
+- Anthropic: default API prompt caching to 1h with configurable TTL override.
+- Anthropic: ignore TTL for OAuth.
+- Auth profiles: keep auto-pinned preference while allowing rotation on failover. (#1138)
+- Auth profiles: user pins stay locked. (#1138)
+- Model catalog: avoid caching import failures, log transient discovery errors, and keep partial results. (#1332)
+- Tests: stabilize Windows gateway/CLI tests by skipping sidecars, normalizing argv, and extending timeouts.
+- Tests: stabilize plugin SDK resolution and embedded agent timeouts.
+- Windows: install gateway scheduled task as the current user.
+- Windows: show friendly guidance instead of failing on access denied.
+- macOS: load menu session previews asynchronously so items populate while the menu is open.
+- macOS: use label colors for session preview text so previews render in menu subviews.
+- macOS: suppress usage error text in the menubar cost view.
+- macOS: Doctor repairs LaunchAgent bootstrap issues for Gateway + Node when listed but not loaded. (#1166)
+- macOS: avoid touching launchd in Remote over SSH so quitting the app no longer disables the remote gateway. (#1105)
+- macOS: bundle Textual resources in packaged app builds to avoid code block crashes. (#1006)
+- Daemon: include HOME in service environments to avoid missing HOME errors. (#1214)
+
+Thanks @AlexMikhalev, @CoreyH, @John-Rood, @KrauseFx, @MaudeBot, @Nachx639, @NicholaiVogel, @RyanLisse, @ThePickle31, @VACInc, @Whoaa512, @YuriNachos, @aaronveklabs, @abdaraxus, @alauppe, @ameno-, @artuskg, @austinm911, @bradleypriest, @cheeeee, @dougvk, @fogboots, @gnarco, @gumadeiras, @jdrhyne, @joelklabo, @longmaba, @mukhtharcm, @odysseus0, @oscargavin, @rhjoh, @sebslight, @sibbl, @sleontenko, @steipete, @suminhthanh, @thewilloftheshadow, @tyler6204, @vignesh07, @visionik, @ysqander, @zerone0x.

 ## 2026.1.16-2

@@ -471,11 +471,6 @@ by Peter Steinberger and the community.
 See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines, maintainers, and how to submit PRs.  
 AI/vibe-coded PRs welcome! 🤖

-Special thanks to @andrewting19 for the Anthropic OAuth tool-name fix.
-
-Core contributors:
- @cpojer — Telegram onboarding UX + docs
-
 Thanks to all clawtributors:

 <p align="left">
@@ -2,6 +2,196 @@
 <rss xmlns:sparkle="http://www.andymatuschak.org/xml-namespaces/sparkle" version="2.0">
    <channel>
        <title>Clawdbot</title>
+        <item>
+            <title>2026.1.20</title>
+            <pubDate>Wed, 21 Jan 2026 08:18:22 +0000</pubDate>
+            <link>https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml</link>
+            <sparkle:version>7116</sparkle:version>
+            <sparkle:shortVersionString>2026.1.20</sparkle:shortVersionString>
+            <sparkle:minimumSystemVersion>15.0</sparkle:minimumSystemVersion>
+            <description><![CDATA[<h2>Clawdbot 2026.1.20</h2>
+<h3>Changes</h3>
+<ul>
+<li>Control UI: add copy-as-markdown with error feedback. (#1345) https://docs.clawd.bot/web/control-ui</li>
+<li>Control UI: drop the legacy list view. (#1345) https://docs.clawd.bot/web/control-ui</li>
+<li>TUI: add syntax highlighting for code blocks. (#1200) https://docs.clawd.bot/tui</li>
+<li>TUI: session picker shows derived titles, fuzzy search, relative times, and last message preview. (#1271) https://docs.clawd.bot/tui</li>
+<li>TUI: add a searchable model picker for quicker model selection. (#1198) https://docs.clawd.bot/tui</li>
+<li>TUI: add input history (up/down) for submitted messages. (#1348) https://docs.clawd.bot/tui</li>
+<li>ACP: add <code>clawdbot acp</code> for IDE integrations. https://docs.clawd.bot/cli/acp</li>
+<li>ACP: add <code>clawdbot acp client</code> interactive harness for debugging. https://docs.clawd.bot/cli/acp</li>
+<li>Skills: add download installs with OS-filtered options. https://docs.clawd.bot/tools/skills</li>
+<li>Skills: add the local sherpa-onnx-tts skill. https://docs.clawd.bot/tools/skills</li>
+<li>Memory: add hybrid BM25 + vector search (FTS5) with weighted merging and fallback. https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: add SQLite embedding cache to speed up reindexing and frequent updates. https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: add OpenAI batch indexing for embeddings when configured. https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: enable OpenAI batch indexing by default for OpenAI embeddings. https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: allow parallel OpenAI batch indexing jobs (default concurrency: 2). https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: render progress immediately, color batch statuses in verbose logs, and poll OpenAI batch status every 2s by default. https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: add <code>--verbose</code> logging for memory status + batch indexing details. https://docs.clawd.bot/concepts/memory</li>
+<li>Memory: add native Gemini embeddings provider for memory search. (#1151) https://docs.clawd.bot/concepts/memory</li>
+<li>Browser: allow config defaults for efficient snapshots in the tool/CLI. (#1336) https://docs.clawd.bot/tools/browser</li>
+<li>Nostr: add the Nostr channel plugin with profile management + onboarding defaults. (#1323) https://docs.clawd.bot/channels/nostr</li>
+<li>Matrix: migrate to matrix-bot-sdk with E2EE support, location handling, and group allowlist upgrades. (#1298) https://docs.clawd.bot/channels/matrix</li>
+<li>Slack: add HTTP webhook mode via Bolt HTTP receiver. (#1143) https://docs.clawd.bot/channels/slack</li>
+<li>Telegram: enrich forwarded-message context with normalized origin details + legacy fallback. (#1090) https://docs.clawd.bot/channels/telegram</li>
+<li>Discord: fall back to <code>/skill</code> when native command limits are exceeded. (#1287)</li>
+<li>Discord: expose <code>/skill</code> globally. (#1287)</li>
+<li>Zalouser: add channel dock metadata, config schema, setup wiring, probe, and status issues. (#1219) https://docs.clawd.bot/plugins/zalouser</li>
+<li>Plugins: require manifest-embedded config schemas with preflight validation warnings. (#1272) https://docs.clawd.bot/plugins/manifest</li>
+<li>Plugins: move channel catalog metadata into plugin manifests. (#1290) https://docs.clawd.bot/plugins/manifest</li>
+<li>Plugins: align Nextcloud Talk policy helpers with core patterns. (#1290) https://docs.clawd.bot/plugins/manifest</li>
+<li>Plugins/UI: let channel plugin metadata drive UI labels/icons and cron channel options. (#1306) https://docs.clawd.bot/web/control-ui</li>
+<li>Plugins: add plugin slots with a dedicated memory slot selector. https://docs.clawd.bot/plugins/agent-tools</li>
+<li>Plugins: ship the bundled BlueBubbles channel plugin (disabled by default). https://docs.clawd.bot/channels/bluebubbles</li>
+<li>Plugins: migrate bundled messaging extensions to the plugin SDK and resolve plugin-sdk imports in the loader.</li>
+<li>Plugins: migrate the Zalo plugin to the shared plugin SDK runtime. https://docs.clawd.bot/channels/zalo</li>
+<li>Plugins: migrate the Zalo Personal plugin to the shared plugin SDK runtime. https://docs.clawd.bot/plugins/zalouser</li>
+<li>Plugins: allow optional agent tools with explicit allowlists and add the plugin tool authoring guide. https://docs.clawd.bot/plugins/agent-tools</li>
+<li>Plugins: auto-enable bundled channel/provider plugins when configuration is present.</li>
+<li>Plugins: sync plugin sources on channel switches and update npm-installed plugins during <code>clawdbot update</code>.</li>
+<li>Plugins: share npm plugin update logic between <code>clawdbot update</code> and <code>clawdbot plugins update</code>.</li>
+<li>Gateway/API: add <code>/v1/responses</code> (OpenResponses) with item-based input + semantic streaming events. (#1229)</li>
+<li>Gateway/API: expand <code>/v1/responses</code> to support file/image inputs, tool_choice, usage, and output limits. (#1229)</li>
+<li>Usage: add <code>/usage cost</code> summaries and macOS menu cost charts. https://docs.clawd.bot/reference/api-usage-costs</li>
+<li>Security: warn when <=300B models run without sandboxing while web tools are enabled. https://docs.clawd.bot/cli/security</li>
+<li>Exec: add host/security/ask routing for gateway + node exec. https://docs.clawd.bot/tools/exec</li>
+<li>Exec: add <code>/exec</code> directive for per-session exec defaults (host/security/ask/node). https://docs.clawd.bot/tools/exec</li>
+<li>Exec approvals: migrate approvals to <code>~/.clawdbot/exec-approvals.json</code> with per-agent allowlists + skill auto-allow toggle, and add approvals UI + node exec lifecycle events. https://docs.clawd.bot/tools/exec-approvals</li>
+<li>Nodes: add headless node host (<code>clawdbot node start</code>) for <code>system.run</code>/<code>system.which</code>. https://docs.clawd.bot/cli/node</li>
+<li>Nodes: add node daemon service install/status/start/stop/restart. https://docs.clawd.bot/cli/node</li>
+<li>Bridge: add <code>skills.bins</code> RPC to support node host auto-allow skill bins.</li>
+<li>Sessions: add daily reset policy with per-type overrides and idle windows (default 4am local), preserving legacy idle-only configs. (#1146) https://docs.clawd.bot/concepts/session</li>
+<li>Sessions: allow <code>sessions_spawn</code> to override thinking level for sub-agent runs. https://docs.clawd.bot/tools/subagents</li>
+<li>Channels: unify thread/topic allowlist matching + command/mention gating helpers across core providers. https://docs.clawd.bot/concepts/groups</li>
+<li>Models: add Qwen Portal OAuth provider support. (#1120) https://docs.clawd.bot/providers/qwen</li>
+<li>Onboarding: add allowlist prompts and username-to-id resolution across core and extension channels. https://docs.clawd.bot/start/onboarding</li>
+<li>Docs: clarify allowlist input types and onboarding behavior for messaging channels. https://docs.clawd.bot/start/onboarding</li>
+<li>Docs: refresh Android node discovery docs for the Gateway WS service type. https://docs.clawd.bot/platforms/android</li>
+<li>Docs: surface Amazon Bedrock in provider lists and clarify Bedrock auth env vars. (#1289) https://docs.clawd.bot/bedrock</li>
+<li>Docs: clarify WhatsApp voice notes. https://docs.clawd.bot/channels/whatsapp</li>
+<li>Docs: clarify Windows WSL portproxy LAN access notes. https://docs.clawd.bot/platforms/windows</li>
+<li>Docs: refresh bird skill install metadata and usage notes. (#1302) https://docs.clawd.bot/tools/browser-login</li>
+<li>Agents: add local docs path resolution and include docs/mirror/source/community pointers in the system prompt.</li>
+<li>Agents: clarify node_modules read-only guidance in agent instructions.</li>
+<li>Config: stamp last-touched metadata on write and warn if the config is newer than the running build.</li>
+<li>macOS: hide usage section when usage is unavailable instead of showing provider errors.</li>
+<li>Android: migrate node transport to the Gateway WebSocket protocol with TLS pinning support + gateway discovery naming.</li>
+<li>Android: send structured payloads in node events/invokes and include user-agent metadata in gateway connects.</li>
+<li>Android: remove legacy bridge transport code now that nodes use the gateway protocol.</li>
+<li>Android: bump okhttp + dnsjava to satisfy lint dependency checks.</li>
+<li>Build: update workspace + core/plugin deps.</li>
+<li>Build: use tsgo for dev/watch builds by default (opt out with <code>CLAWDBOT_TS_COMPILER=tsc</code>).</li>
+<li>Repo: remove the Peekaboo git submodule now that the SPM release is used.</li>
+<li>macOS: switch PeekabooBridge integration to the tagged Swift Package Manager release.</li>
+<li>macOS: stop syncing Peekaboo in postinstall.</li>
+<li>Swabble: use the tagged Commander Swift package release.</li>
+</ul>
+<h3>Breaking</h3>
+<ul>
+<li><strong>BREAKING:</strong> Reject invalid/unknown config entries and refuse to start the gateway for safety. Run <code>clawdbot doctor --fix</code> to repair, then update plugins (<code>clawdbot plugins update</code>) if you use any.</li>
+</ul>
+<h3>Fixes</h3>
+<ul>
+<li>Discovery: shorten Bonjour DNS-SD service type to <code>_clawdbot-gw._tcp</code> and update discovery clients/docs.</li>
+<li>Diagnostics: export OTLP logs, correct queue depth tracking, and document message-flow telemetry.</li>
+<li>Diagnostics: emit message-flow diagnostics across channels via shared dispatch. (#1244)</li>
+<li>Diagnostics: gate heartbeat/webhook logging. (#1244)</li>
+<li>Gateway: strip inbound envelope headers from chat history messages to keep clients clean.</li>
+<li>Gateway: clarify unauthorized handshake responses with token/password mismatch guidance.</li>
+<li>Gateway: allow mobile node client ids for iOS + Android handshake validation. (#1354)</li>
+<li>Gateway: clarify connect/validation errors for gateway params. (#1347)</li>
+<li>Gateway: preserve restart wake routing + thread replies across restarts. (#1337)</li>
+<li>Gateway: reschedule per-agent heartbeats on config hot reload without restarting the runner.</li>
+<li>Gateway: require authorized restarts for SIGUSR1 (restart/apply/update) so config gating can't be bypassed.</li>
+<li>Cron: auto-deliver isolated agent output to explicit targets without tool calls. (#1285)</li>
+<li>Agents: preserve subagent announce thread/topic routing + queued replies across channels. (#1241)</li>
+<li>Agents: propagate accountId into embedded runs so sub-agent announce routing honors the originating account. (#1058)</li>
+<li>Agents: avoid treating timeout errors with "aborted" messages as user aborts, so model fallback still runs. (#1137)</li>
+<li>Agents: sanitize oversized image payloads before send and surface image-dimension errors.</li>
+<li>Sessions: fall back to session labels when listing display names. (#1124)</li>
+<li>Compaction: include tool failure summaries in safeguard compaction to prevent retry loops. (#1084)</li>
+<li>Config: log invalid config issues once per run and keep invalid-config errors stackless.</li>
+<li>Config: allow Perplexity as a web_search provider in config validation. (#1230)</li>
+<li>Config: allow custom fields under <code>skills.entries.<name>.config</code> for skill credentials/config. (#1226)</li>
+<li>Doctor: clarify plugin auto-enable hint text in the startup banner.</li>
+<li>Doctor: canonicalize legacy session keys in session stores to prevent stale metadata. (#1169)</li>
+<li>Docs: make docs:list fail fast with a clear error if the docs directory is missing.</li>
+<li>Plugins: add Nextcloud Talk manifest for plugin config validation. (#1297)</li>
+<li>Plugins: surface plugin load/register/config errors in gateway logs with plugin/source context.</li>
+<li>CLI: preserve cron delivery settings when editing message payloads. (#1322)</li>
+<li>CLI: keep <code>clawdbot logs</code> output resilient to broken pipes while preserving progress output.</li>
+<li>CLI: avoid duplicating --profile/--dev flags when formatting commands.</li>
+<li>CLI: centralize CLI command registration to keep fast-path routing and program wiring in sync. (#1207)</li>
+<li>CLI: keep banners on routed commands, restore config guarding outside fast-path routing, and tighten fast-path flag parsing while skipping console capture for extra speed. (#1195)</li>
+<li>CLI: skip runner rebuilds when dist is fresh. (#1231)</li>
+<li>CLI: add WSL2/systemd unavailable hints in daemon status/doctor output.</li>
+<li>Status: route native <code>/status</code> to the active agent so model selection reflects the correct profile. (#1301)</li>
+<li>Status: show both usage windows with reset hints when usage data is available. (#1101)</li>
+<li>UI: keep config form enums typed, preserve empty strings, protect sensitive defaults, and deepen config search. (#1315)</li>
+<li>UI: preserve ordered list numbering in chat markdown. (#1341)</li>
+<li>UI: allow Control UI to read gatewayUrl from URL params for remote WebSocket targets. (#1342)</li>
+<li>UI: prevent double-scroll in Control UI chat by locking chat layout to the viewport. (#1283)</li>
+<li>UI: enable shell mode for sync Windows spawns to avoid <code>pnpm ui:build</code> EINVAL. (#1212)</li>
+<li>TUI: keep thinking blocks ordered before content during streaming and isolate per-run assembly. (#1202)</li>
+<li>TUI: align custom editor initialization with the latest pi-tui API. (#1298)</li>
+<li>TUI: show generic empty-state text for searchable pickers. (#1201)</li>
+<li>TUI: highlight model search matches and stabilize search ordering.</li>
+<li>Configure: hide OpenRouter auto routing model from the model picker. (#1182)</li>
+<li>Memory: show total file counts + scan issues in <code>clawdbot memory status</code>.</li>
+<li>Memory: fall back to non-batch embeddings after repeated batch failures.</li>
+<li>Memory: apply OpenAI batch defaults even without explicit remote config.</li>
+<li>Memory: index atomically so failed reindex preserves the previous memory database. (#1151)</li>
+<li>Memory: avoid sqlite-vec unique constraint failures when reindexing duplicate chunk ids. (#1151)</li>
+<li>Memory: retry transient 5xx errors (Cloudflare) during embedding indexing.</li>
+<li>Memory: parallelize embedding indexing with rate-limit retries.</li>
+<li>Memory: split overly long lines to keep embeddings under token limits.</li>
+<li>Memory: skip empty chunks to avoid invalid embedding inputs.</li>
+<li>Memory: split embedding batches to avoid OpenAI token limits during indexing.</li>
+<li>Memory: probe sqlite-vec availability in <code>clawdbot memory status</code>.</li>
+<li>Exec approvals: enforce allowlist when ask is off.</li>
+<li>Exec approvals: prefer raw command for node approvals/events.</li>
+<li>Tools: show exec elevated flag before the command and keep it outside markdown in tool summaries.</li>
+<li>Tools: return a companion-app-required message when node exec is requested with no paired node.</li>
+<li>Tools: return a companion-app-required message when <code>system.run</code> is requested without a supporting node.</li>
+<li>Exec: default gateway/node exec security to allowlist when unset (sandbox stays deny).</li>
+<li>Exec: prefer bash when fish is default shell, falling back to sh if bash is missing. (#1297)</li>
+<li>Exec: merge login-shell PATH for host=gateway exec while keeping daemon PATH minimal. (#1304)</li>
+<li>Streaming: emit assistant deltas for OpenAI-compatible SSE chunks. (#1147)</li>
+<li>Discord: make resolve warnings avoid raw JSON payloads on rate limits.</li>
+<li>Discord: process message handlers in parallel across sessions to avoid event queue blocking. (#1295)</li>
+<li>Discord: stop reconnecting the gateway after aborts to prevent duplicate listeners.</li>
+<li>Discord: only emit slow listener warnings after 30s.</li>
+<li>Discord: inherit parent channel allowlists for thread slash commands and reactions. (#1123)</li>
+<li>Telegram: honor pairing allowlists for native slash commands.</li>
+<li>Telegram: preserve hidden text_link URLs by expanding entities in inbound text. (#1118)</li>
+<li>Slack: resolve Bolt import interop for Bun + Node. (#1191)</li>
+<li>Web search: infer Perplexity base URL from API key source (direct vs OpenRouter).</li>
+<li>Web fetch: harden SSRF protection with shared hostname checks and redirect limits. (#1346)</li>
+<li>Browser: register AI snapshot refs for act commands. (#1282)</li>
+<li>Voice call: include request query in Twilio webhook verification when publicUrl is set. (#864)</li>
+<li>Anthropic: default API prompt caching to 1h with configurable TTL override.</li>
+<li>Anthropic: ignore TTL for OAuth.</li>
+<li>Auth profiles: keep auto-pinned preference while allowing rotation on failover. (#1138)</li>
+<li>Auth profiles: user pins stay locked. (#1138)</li>
+<li>Model catalog: avoid caching import failures, log transient discovery errors, and keep partial results. (#1332)</li>
+<li>Tests: stabilize Windows gateway/CLI tests by skipping sidecars, normalizing argv, and extending timeouts.</li>
+<li>Tests: stabilize plugin SDK resolution and embedded agent timeouts.</li>
+<li>Windows: install gateway scheduled task as the current user.</li>
+<li>Windows: show friendly guidance instead of failing on access denied.</li>
+<li>macOS: load menu session previews asynchronously so items populate while the menu is open.</li>
+<li>macOS: use label colors for session preview text so previews render in menu subviews.</li>
+<li>macOS: suppress usage error text in the menubar cost view.</li>
+<li>macOS: Doctor repairs LaunchAgent bootstrap issues for Gateway + Node when listed but not loaded. (#1166)</li>
+<li>macOS: avoid touching launchd in Remote over SSH so quitting the app no longer disables the remote gateway. (#1105)</li>
+<li>macOS: bundle Textual resources in packaged app builds to avoid code block crashes. (#1006)</li>
+<li>Daemon: include HOME in service environments to avoid missing HOME errors. (#1214)</li>
+</ul>
+Thanks @AlexMikhalev, @CoreyH, @John-Rood, @KrauseFx, @MaudeBot, @Nachx639, @NicholaiVogel, @RyanLisse, @ThePickle31, @VACInc, @Whoaa512, @YuriNachos, @aaronveklabs, @abdaraxus, @alauppe, @ameno-, @artuskg, @austinm911, @bradleypriest, @cheeeee, @dougvk, @fogboots, @gnarco, @gumadeiras, @jdrhyne, @joelklabo, @longmaba, @mukhtharcm, @odysseus0, @oscargavin, @rhjoh, @sebslight, @sibbl, @sleontenko, @steipete, @suminhthanh, @thewilloftheshadow, @tyler6204, @vignesh07, @visionik, @ysqander, @zerone0x.
+<p><a href="https://github.com/clawdbot/clawdbot/blob/main/CHANGELOG.md">View full changelog</a></p>
+]]></description>
+            <enclosure url="https://github.com/clawdbot/clawdbot/releases/download/v2026.1.20/Clawdbot-2026.1.20.zip" length="12208102" type="application/octet-stream" sparkle:edSignature="hU495Eii8O3qmmUnxYFhXyEGv+qan6KL+GpeuBhPIXf+7B5F/gBh5Oz9cHaqaAPoZ4/3Bo6xgvic0HTkbz6gDw=="/>
+        </item>
        <item>
            <title>2026.1.16-2</title>
            <pubDate>Sat, 17 Jan 2026 12:46:22 +0000</pubDate>
@@ -99,177 +289,5 @@
 ]]></description>
            <enclosure url="https://github.com/clawdbot/clawdbot/releases/download/v2026.1.15/Clawdbot-2026.1.15.zip" length="12127276" type="application/octet-stream" sparkle:edSignature="o79vwTbtW/d91NQFRVfUDhsv6D4zIw7IkhY0N1iLImMu94BURgLcecA6z7Smy3bMobPwOyzN8yfm6mA/Rt8FCA=="/>
        </item>
-        <item>
-            <title>2026.1.14-1</title>
-            <pubDate>Thu, 15 Jan 2026 11:14:40 +0000</pubDate>
-            <link>https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml</link>
-            <sparkle:version>5825</sparkle:version>
-            <sparkle:shortVersionString>2026.1.14-1</sparkle:shortVersionString>
-            <sparkle:minimumSystemVersion>15.0</sparkle:minimumSystemVersion>
-            <description><![CDATA[<h2>Clawdbot 2026.1.14-1</h2>
-<h3>Highlights</h3>
-<ul>
-<li>Web search: <code>web_search</code>/<code>web_fetch</code> tools (Brave API) + first-time setup in onboarding/configure.</li>
-<li>Browser control: Chrome extension relay takeover mode + remote browser control via <code>clawdbot browser serve</code>.</li>
-<li>Plugins: channel plugins (gateway HTTP hooks) + Zalo plugin + onboarding install flow. (#854) — thanks @longmaba.</li>
-<li>Security: expanded <code>clawdbot security audit</code> (+ <code>--fix</code>), detect-secrets CI scan, and a <code>SECURITY.md</code> reporting policy.</li>
-</ul>
-<h3>Changes</h3>
-<h4>Web Tools</h4>
-<ul>
-<li>Tools: add <code>web_search</code>/<code>web_fetch</code> (Brave API), including helpful setup hints when the key is missing.</li>
-<li>Tools: enable <code>web_fetch</code> by default (unless explicitly disabled in config).</li>
-<li>CLI/Docs: add <code>clawdbot configure --section web</code> for storing Brave API keys and update onboarding tips.</li>
-</ul>
-<h4>Browser / Control UI</h4>
-<ul>
-<li>Browser: add Chrome extension relay takeover mode (toolbar button) + <code>clawdbot browser serve</code> remote control + <code>browser.controlToken</code>.</li>
-<li>Browser: ship a built-in <code>chrome</code> profile for extension relay and start the relay automatically when running locally.</li>
-<li>Browser: default <code>browser.defaultProfile</code> to <code>chrome</code> (existing Chrome takeover mode).</li>
-<li>Browser: add <code>clawdbot browser extension install/path</code> and copy extension path to clipboard.</li>
-<li>Browser: add <code>snapshot refs=aria</code> (Playwright aria-ref ids) for self-resolving refs across <code>snapshot</code> → <code>act</code>.</li>
-<li>Browser: <code>profile="chrome"</code> now defaults to host control and returns clearer “attach a tab” errors.</li>
-<li>Browser: extension mode recovers when only one tab is attached (stale targetId fallback).</li>
-<li>Control UI: show raw any-map entries in config views; move Docs link into the left nav.</li>
-</ul>
-<h4>Plugins</h4>
-<ul>
-<li>Plugins: add plugin HTTP hooks + loader updates to support channel plugins. (#854) — thanks @longmaba.</li>
-<li>Plugins: add onboarding plugin install flow. (#854) — thanks @longmaba.</li>
-<li>Channels: add Matrix plugin (external) with docs + onboarding hooks.</li>
-<li>Voice Call: add Plivo provider (no SDK dependency). (#846) — thanks @vrknetha.</li>
-</ul>
-<h4>Security</h4>
-<ul>
-<li>Security: expand <code>clawdbot security audit</code> checks and publish a <code>SECURITY.md</code> reporting policy.</li>
-<li>Security: extend <code>clawdbot security audit --fix</code> to tighten more sensitive state paths.</li>
-<li>Security: add detect-secrets CI scan and baseline guidance. (#227) — thanks @Hyaxia.</li>
-</ul>
-<h4>Onboarding / Daemon</h4>
-<ul>
-<li>Onboarding: add a security checkpoint prompt (docs link + sandboxing hint); require <code>--accept-risk</code> for <code>--non-interactive</code>.</li>
-<li>Daemon: support profile-aware service names for multi-gateway setups. (#671) — thanks @bjesuiter.</li>
-</ul>
-<h4>Auth / Usage / Config</h4>
-<ul>
-<li>Usage: add MiniMax coding plan usage tracking.</li>
-<li>Auth: label Claude Code CLI auth options. (#915) — thanks @SeanZoR.</li>
-<li>Agents: add optional auth-profile copy prompt on <code>agents add</code> and improve auth error messaging.</li>
-<li>Auth: add dynamic template variables to <code>messages.responsePrefix</code>. (#928) — thanks @sebslight.</li>
-<li>Config: add <code>channels.<provider>.configWrites</code> gating for channel-initiated config writes; migrate Slack channel IDs.</li>
-</ul>
-<h4>Channels</h4>
-<ul>
-<li>Telegram: add message delete action in the message tool. (#903) — thanks @sleontenko.</li>
-<li>WhatsApp: add <code>channels.whatsapp.sendReadReceipts</code> to disable auto read receipts. (#882) — thanks @chrisrodz.</li>
-</ul>
-<h4>Docs</h4>
-<ul>
-<li>Docs: clarify per-agent auth stores, sandboxed skill binaries, and elevated semantics.</li>
-<li>Docs: add FAQ entries for missing provider auth after adding agents and Gemini thinking signature errors.</li>
-<li>Docs: expand gateway security hardening guidance and incident response checklist.</li>
-<li>Docs: document DM history limits for channel DMs. (#883) — thanks @pkrmf.</li>
-<li>Docs: standardize Claude Code CLI naming across docs and prompts. (follow-up to #915)</li>
-<li>Docs: add per-command CLI doc pages and link them from <code>clawdbot <command> --help</code>.</li>
-<li>Docs: add multi-gateway guide (sidebar + nav).</li>
-</ul>
-<h3>Fixes</h3>
-<h4>Gateway / Daemon / Sessions</h4>
-<ul>
-<li>Gateway: forward termination signals to respawned CLI child processes to avoid orphaned systemd runs. (#933) — thanks @roshanasingh4.</li>
-<li>Gateway/UI: ship session defaults in the hello snapshot so the Control UI canonicalizes main session keys (no bare <code>main</code> alias).</li>
-<li>Agents: skip thinking/final tag stripping inside Markdown code spans. (#939) — thanks @ngutman.</li>
-<li>Browser: add tests for snapshot labels/efficient query params and labeled image responses.</li>
-<li>Browser: persist role snapshot refs per CDP target so <code>snapshot</code> → <code>act</code> clicks work even if Playwright returns a different Page instance.</li>
-<li>macOS: ensure launchd log directory exists with a test-only override. (#909) — thanks @roshanasingh4.</li>
-<li>macOS: format ConnectionsStore config to satisfy SwiftFormat lint. (#852) — thanks @mneves75.</li>
-<li>Packaging: run <code>pnpm build</code> on <code>prepack</code> so npm publishes include fresh <code>dist/</code> output.</li>
-<li>Telegram: register dock native commands with underscores to avoid <code>BOT_COMMAND_INVALID</code> (#929, fixes #901) — thanks @grp06.</li>
-<li>Google: downgrade unsigned thinking blocks before send to avoid missing signature errors.</li>
-<li>Agents: make user time zone and 24-hour time explicit in the system prompt. (#859) — thanks @CashWilliams.</li>
-<li>Agents: strip downgraded tool call text without eating adjacent replies and filter thinking-tag leaks. (#905) — thanks @erikpr1994.</li>
-<li>Agents: cap tool call IDs for OpenAI/OpenRouter to avoid request rejections. (#875) — thanks @j1philli.</li>
-<li>Doctor: avoid re-adding WhatsApp config when only legacy ack reactions are set. (#927, fixes #900) — thanks @grp06.</li>
-<li>Agents: scrub tuple <code>items</code> schemas for Gemini tool calls. (#926, fixes #746) — thanks @grp06.</li>
-<li>Agents: stabilize sub-agent announce status from runtime outcomes and normalize Result/Notes. (#835) — thanks @roshanasingh4.</li>
-<li>Apps: use canonical main session keys from gateway defaults across macOS/iOS/Android to avoid creating bare <code>main</code> sessions.</li>
-<li>Embedded runner: suppress raw API error payloads from replies. (#924) — thanks @grp06.</li>
-<li>Auth: normalize Claude Code CLI profile mode to oauth and auto-migrate config. (#855) — thanks @sebslight.</li>
-<li>Daemon: clear persisted launchd disabled state before bootstrap (fixes <code>daemon install</code> after uninstall). (#849) — thanks @ndraiman.</li>
-<li>Sessions: return deep clones (<code>structuredClone</code>) so cached session entries can't be mutated. (#934) — thanks @ronak-guliani.</li>
-<li>Heartbeat: keep <code>updatedAt</code> monotonic when restoring heartbeat sessions. (#934) — thanks @ronak-guliani.</li>
-<li>Agent: clear run context after CLI runs (<code>clearAgentRunContext</code>) to avoid runaway contexts. (#934) — thanks @ronak-guliani.</li>
-<li>Gateway/Dev: ensure <code>pnpm gateway:dev</code> always uses the dev profile config + state (<code>~/.clawdbot-dev</code>).</li>
-</ul>
-<h4>CLI / Onboarding</h4>
-<ul>
-<li>Onboarding: show web search setup at the end (not the beginning).</li>
-<li>Onboarding: show daemon install/restart progress (avoid “blinking cursor”) and fix daemon install output formatting.</li>
-<li>Health: colorize “not configured” provider lines for easier scanning.</li>
-</ul>
-<h4>Control UI / TUI</h4>
-<ul>
-<li>Control UI: load cron run history on job selection and clarify empty-state messaging. (#866)</li>
-<li>UI: use application-defined WebSocket close code and fix dashboard auth query items. (#918) — thanks @rahthakor.</li>
-<li>UI: always apply <code>?token=</code> from URL (fixes unauthorized after re-onboard).</li>
-<li>Browser: add tests for snapshot labels/efficient query params and labeled image responses.</li>
-<li>TUI: render picker overlays via the overlay stack so /models and /settings display. (#921) — thanks @grizzdank.</li>
-<li>TUI: add a bright spinner + elapsed time in the status line for send/stream/run states.</li>
-<li>TUI: show LLM error messages (rate limits, auth, etc.) instead of <code>(no output)</code>.</li>
-</ul>
-<h4>Agents / Auth / Tools / Sandbox</h4>
-<ul>
-<li>Agents: make user time zone and 24-hour time explicit in the system prompt. (#859) — thanks @CashWilliams.</li>
-<li>Agents: strip downgraded tool call text without eating adjacent replies and filter thinking-tag leaks. (#905) — thanks @erikpr1994.</li>
-<li>Agents: cap tool call IDs for OpenAI/OpenRouter to avoid request rejections. (#875) — thanks @j1philli.</li>
-<li>Agents: scrub tuple <code>items</code> schemas for Gemini tool calls. (#926, fixes #746) — thanks @grp06.</li>
-<li>Agents: stabilize sub-agent announce status from runtime outcomes and normalize Result/Notes. (#835) — thanks @roshanasingh4.</li>
-<li>Auth: normalize Claude Code CLI profile mode to oauth and auto-migrate config. (#855) — thanks @sebslight.</li>
-<li>Embedded runner: suppress raw API error payloads from replies. (#924) — thanks @grp06.</li>
-<li>Logging: tolerate <code>EIO</code> from console writes to avoid gateway crashes. (#925, fixes #878) — thanks @grp06.</li>
-<li>Sandbox: restore <code>docker.binds</code> config validation and preserve configured PATH for <code>docker exec</code>. (#873) — thanks @akonyer.</li>
-<li>Google: downgrade unsigned thinking blocks before send to avoid missing signature errors.</li>
-</ul>
-<h4>macOS / Apps</h4>
-<ul>
-<li>macOS: ensure launchd log directory exists with a test-only override. (#909) — thanks @roshanasingh4.</li>
-<li>macOS: format ConnectionsStore config to satisfy SwiftFormat lint. (#852) — thanks @mneves75.</li>
-<li>macOS: pass auth token/password to dashboard URL for authenticated access. (#918) — thanks @rahthakor.</li>
-<li>macOS: reuse launchd gateway auth and skip wizard when gateway config already exists. (#917)</li>
-<li>Apps: use canonical main session keys from gateway defaults across macOS/iOS/Android to avoid creating bare <code>main</code> sessions.</li>
-<li>macOS: fix cron preview/testing payload to use <code>channel</code> key. (#867) — thanks @wes-davis.</li>
-<li>macOS: update cron testing channel arg. (#896) — thanks @ngutman.</li>
-</ul>
-<h4>Channels / Messaging</h4>
-<ul>
-<li>Slack: isolate thread history and avoid inheriting channel transcripts for new threads by default. (#758)</li>
-<li>Slack: respect <code>channels.slack.requireMention</code> default when resolving channel mention gating. (#850) — thanks @evalexpr.</li>
-<li>Slack: drop Socket Mode events with mismatched <code>api_app_id</code>/<code>team_id</code>. (#889) — thanks @roshanasingh4.</li>
-<li>Commands: add native command argument menus across Discord/Slack/Telegram. (#936) — thanks @thewilloftheshadow.</li>
-<li>Discord: isolate autoThread thread context. (#856) — thanks @davidguttman.</li>
-<li>Telegram: honor <code>channels.telegram.timeoutSeconds</code> for grammY API requests. (#863) — thanks @Snaver.</li>
-<li>Telegram: aggregate split inbound messages into one prompt (reduces “one reply per fragment”).</li>
-<li>Telegram: let control commands bypass per-chat sequentialization; always allow abort triggers.</li>
-<li>Telegram: split long captions into media + follow-up text messages. (#907) — thanks @jalehman.</li>
-<li>Telegram: migrate group config when supergroups change chat IDs. (#906) — thanks @sleontenko.</li>
-<li>Telegram: register dock native commands with underscores to avoid <code>BOT_COMMAND_INVALID</code> (#929, fixes #901) — thanks @grp06.</li>
-<li>Messaging: unify markdown formatting + format-first chunking for Slack/Telegram/Signal. (#920) — thanks @TheSethRose.</li>
-<li>iMessage: prefer handle routing for direct-message replies; include imsg RPC error details. (#935)</li>
-<li>WhatsApp: fix context isolation using wrong ID (was bot's number, now conversation ID). (#911) — thanks @tristanmanchester.</li>
-<li>WhatsApp: normalize user JIDs with device suffix for allowlist checks in groups. (#838) — thanks @peschee.</li>
-<li>WhatsApp: harden owner command auth.</li>
-<li>Auto-reply: treat trailing <code>NO_REPLY</code> tokens as silent replies.</li>
-</ul>
-<h4>Config / Doctor / Packaging</h4>
-<ul>
-<li>Config: prevent partial config writes from clobbering unrelated settings (base hash guard + merge patch for connection saves).</li>
-<li>Config/Doctor: remove legacy Clawdis env fallbacks and config/service migrations (Clawdbot-only).</li>
-<li>Doctor: avoid re-adding WhatsApp config when only legacy ack reactions are set. (#927, fixes #900) — thanks @grp06.</li>
-<li>Packaging: run <code>pnpm build</code> on <code>prepack</code> so npm publishes include fresh <code>dist/</code> output.</li>
-</ul>
-<p><a href="https://github.com/clawdbot/clawdbot/blob/main/CHANGELOG.md">View full changelog</a></p>
-]]></description>
-            <enclosure url="https://github.com/clawdbot/clawdbot/releases/download/v2026.1.14-1/Clawdbot-2026.1.14-1.zip" length="19887144" type="application/octet-stream" sparkle:edSignature="1irKxBLt2eRtns34m/8JsjL/ZzhZQNjahwrxtArTvzaCnidS/MEnpD4nV2SHnhuo8g+fJZQpV9NoCAoEOAinCw=="/>
-        </item>
    </channel>
 </rss>
@@ -276,12 +276,13 @@ enum ExecApprovalsStore {
            ? agentId!.trimmingCharacters(in: .whitespacesAndNewlines)
            : "default"
        let agentEntry = file.agents?[key] ?? ExecApprovalsAgent()
+        let wildcardEntry = file.agents?["*"] ?? ExecApprovalsAgent()
        let resolvedAgent = ExecApprovalsResolvedDefaults(
-            security: agentEntry.security ?? resolvedDefaults.security,
-            ask: agentEntry.ask ?? resolvedDefaults.ask,
-            askFallback: agentEntry.askFallback ?? resolvedDefaults.askFallback,
-            autoAllowSkills: agentEntry.autoAllowSkills ?? resolvedDefaults.autoAllowSkills)
-        let allowlist = (agentEntry.allowlist ?? [])
+            security: agentEntry.security ?? wildcardEntry.security ?? resolvedDefaults.security,
+            ask: agentEntry.ask ?? wildcardEntry.ask ?? resolvedDefaults.ask,
+            askFallback: agentEntry.askFallback ?? wildcardEntry.askFallback ?? resolvedDefaults.askFallback,
+            autoAllowSkills: agentEntry.autoAllowSkills ?? wildcardEntry.autoAllowSkills ?? resolvedDefaults.autoAllowSkills)
+        let allowlist = ((wildcardEntry.allowlist ?? []) + (agentEntry.allowlist ?? []))
            .map { entry in
                ExecAllowlistEntry(
                    pattern: entry.pattern.trimmingCharacters(in: .whitespacesAndNewlines),
@@ -178,7 +178,7 @@ final class TailscaleService {
        }
    }

-    private static func isTailnetIPv4(_ address: String) -> Bool {
+    private nonisolated static func isTailnetIPv4(_ address: String) -> Bool {
        let parts = address.split(separator: ".")
        guard parts.count == 4 else { return false }
        let octets = parts.compactMap { Int($0) }
@@ -188,7 +188,7 @@ final class TailscaleService {
        return a == 100 && b >= 64 && b <= 127
    }

-    private static func detectTailnetIPv4() -> String? {
+    private nonisolated static func detectTailnetIPv4() -> String? {
        var addrList: UnsafeMutablePointer<ifaddrs>?
        guard getifaddrs(&addrList) == 0, let first = addrList else { return nil }
        defer { freeifaddrs(addrList) }
@@ -7,8 +7,8 @@ read_when:

 # `clawdbot approvals`

-Manage exec approvals for the **gateway host** or a **node host**.
-By default, commands target the gateway. Use `--node` to edit a node’s approvals.
+Manage exec approvals for the **local host**, **gateway host**, or a **node host**.
+By default, commands target the local approvals file on disk. Use `--gateway` to target the gateway, or `--node` to target a specific node.

 Related:
 - Exec approvals: [Exec approvals](/tools/exec-approvals)
@@ -19,6 +19,7 @@ Related:
 ```bash
 clawdbot approvals get
 clawdbot approvals get --node <id|name|ip>
+clawdbot approvals get --gateway
 ```

 ## Replace approvals from a file
@@ -26,6 +27,7 @@ clawdbot approvals get --node <id|name|ip>
 ```bash
 clawdbot approvals set --file ./exec-approvals.json
 clawdbot approvals set --node <id|name|ip> --file ./exec-approvals.json
+clawdbot approvals set --gateway --file ./exec-approvals.json
 ```

 ## Allowlist helpers
@@ -33,6 +35,7 @@ clawdbot approvals set --node <id|name|ip> --file ./exec-approvals.json
 ```bash
 clawdbot approvals allowlist add "~/Projects/**/bin/rg"
 clawdbot approvals allowlist add --agent main --node <id|name|ip> "/usr/bin/uptime"
+clawdbot approvals allowlist add --agent "*" "/usr/bin/uname"

 clawdbot approvals allowlist remove "~/Projects/**/bin/rg"
 ```
@@ -40,5 +43,6 @@ clawdbot approvals allowlist remove "~/Projects/**/bin/rg"
 ## Notes

 - `--node` uses the same resolver as `clawdbot nodes` (id, name, ip, or id prefix).
+- `--agent` defaults to `"*"`, which applies to all agents.
 - The node host must advertise `system.execApprovals.get/set` (macOS app or headless node host).
 - Approvals files are stored per host at `~/.clawdbot/exec-approvals.json`.
@@ -8,6 +8,9 @@ read_when:

 Interactive prompt to set up credentials, devices, and agent defaults.

+Note: The **Model** section now includes a multi-select for the
+`agents.defaults.models` allowlist (what shows up in `/model` and the model picker).
+
 Tip: `clawdbot config` without a subcommand opens the same wizard. Use
 `clawdbot config get|set|unset` for non-interactive edits.

@@ -9,7 +9,7 @@ read_when:

 Manage the Gateway daemon (background service).

-Note: `clawdbot service gateway …` is the preferred surface; `daemon` remains
+Note: `clawdbot daemon …` is the preferred surface for Gateway service management; `daemon` remains
 as a legacy alias for compatibility.

 Related:
@@ -29,7 +29,6 @@ This page describes the current CLI behavior. If commands change, update this do
 - [`sessions`](/cli/sessions)
 - [`gateway`](/cli/gateway)
 - [`daemon`](/cli/daemon)
- [`service`](/cli/service)
 - [`logs`](/cli/logs)
 - [`models`](/cli/models)
 - [`memory`](/cli/memory)
@@ -146,21 +145,6 @@ clawdbot [--dev] [--profile <name>] <command>
    start
    stop
    restart
-  service
-    gateway
-      status
-      install
-      uninstall
-      start
-      stop
-      restart
-    node
-      status
-      install
-      uninstall
-      start
-      stop
-      restart
  logs
  models
    list
@@ -544,7 +528,7 @@ Options:
 - `--debug` (alias for `--verbose`)

 Notes:
- Overview includes Gateway + Node service status when available.
+- Overview includes Gateway + node host service status when available.

 ### Usage tracking
 Clawdbot can surface provider usage/quota when OAuth/API creds are available.
@@ -806,16 +790,13 @@ All `cron` commands accept `--url`, `--token`, `--timeout`, `--expect-final`.
 [`clawdbot node`](/cli/node).

 Subcommands:
- `node start --host <gateway-host> --port 18790`
- `node service status`
- `node service install [--host <gateway-host>] [--port <port>] [--tls] [--tls-fingerprint <sha256>] [--node-id <id>] [--display-name <name>] [--runtime <node|bun>] [--force]`
- `node service uninstall`
- `node service start`
- `node service stop`
- `node service restart`
-
-Legacy alias:
- `node daemon …` (same as `node service …`)
+- `node run --host <gateway-host> --port 18790`
+- `node status`
+- `node install [--host <gateway-host>] [--port <port>] [--tls] [--tls-fingerprint <sha256>] [--node-id <id>] [--display-name <name>] [--runtime <node|bun>] [--force]`
+- `node uninstall`
+- `node run`
+- `node stop`
+- `node restart`

 ## Nodes

@@ -23,10 +23,10 @@ Common use cases:
 Execution is still guarded by **exec approvals** and per‑agent allowlists on the
 node host, so you can keep command access scoped and explicit.

-## Start (foreground)
+## Run (foreground)

 ```bash
-clawdbot node start --host <gateway-host> --port 18790
+clawdbot node run --host <gateway-host> --port 18790
 ```

 Options:
@@ -42,9 +42,7 @@ Options:
 Install a headless node host as a user service.

 ```bash
-clawdbot node service install --host <gateway-host> --port 18790
-# or
-clawdbot service node install --host <gateway-host> --port 18790
+clawdbot node install --host <gateway-host> --port 18790
 ```

 Options:
@@ -61,18 +59,10 @@ Manage the service:

 ```bash
 clawdbot node status
-clawdbot service node status
-clawdbot node service status
-clawdbot node service start
-clawdbot node service stop
-clawdbot node service restart
-clawdbot node service uninstall
-```
-
-Legacy alias:
-
-```bash
-clawdbot node daemon status
+clawdbot node run
+clawdbot node stop
+clawdbot node restart
+clawdbot node uninstall
 ```

 ## Pairing
@@ -1,51 +0,0 @@
---
-summary: "CLI reference for `clawdbot service` (manage gateway + node services)"
-read_when:
-  - You want to manage Gateway or node services cross-platform
-  - You want a single surface for start/stop/install/uninstall
---
-
-# `clawdbot service`
-
-Manage the **Gateway** service and **node host** services.
-
-Related:
- Gateway daemon (legacy alias): [Daemon](/cli/daemon)
- Node host: [Node](/cli/node)
-
-## Gateway service
-
-```bash
-clawdbot service gateway status
-clawdbot service gateway install --port 18789
-clawdbot service gateway start
-clawdbot service gateway stop
-clawdbot service gateway restart
-clawdbot service gateway uninstall
-```
-
-Notes:
- `service gateway status` supports `--json` and `--deep` for system checks.
- `service gateway install` supports `--runtime node|bun` and `--token`.
-
-## Node host service
-
-```bash
-clawdbot service node status
-clawdbot service node install --host <gateway-host> --port 18790
-clawdbot service node start
-clawdbot service node stop
-clawdbot service node restart
-clawdbot service node uninstall
-```
-
-Notes:
- `service node install` supports `--runtime node|bun`, `--node-id`, `--display-name`,
-  and TLS options (`--tls`, `--tls-fingerprint`).
-
-## Aliases
-
- `clawdbot daemon …` → `clawdbot service gateway …`
- `clawdbot node service …` → `clawdbot service node …`
- `clawdbot node status` → `clawdbot service node status`
- `clawdbot node daemon …` → `clawdbot service node …` (legacy)
@@ -19,6 +19,6 @@ clawdbot status --usage
 Notes:
 - `--deep` runs live probes (WhatsApp Web + Telegram + Discord + Slack + Signal).
 - Output includes per-agent session stores when multiple agents are configured.
- Overview includes Gateway + Node service install/runtime status when available.
+- Overview includes Gateway + node host service install/runtime status when available.
 - Overview includes update channel + git SHA (for source checkouts).
 - Update info surfaces in the Overview; if an update is available, status prints a hint to run `clawdbot update` (see [Updating](/install/updating)).
@@ -194,7 +194,7 @@ Local mode:

 - File type: Markdown only (`MEMORY.md`, `memory/**/*.md`).
 - Index storage: per-agent SQLite at `~/.clawdbot/memory/<agentId>.sqlite` (configurable via `agents.defaults.memorySearch.store.path`, supports `{agentId}` token).
- Freshness: watcher on `MEMORY.md` + `memory/` marks the index dirty (debounce 1.5s). Sync runs on session start, on first search when dirty, and optionally on an interval.
+- Freshness: watcher on `MEMORY.md` + `memory/` marks the index dirty (debounce 1.5s). Sync is scheduled on session start, on search, or on an interval and runs asynchronously. Session transcripts use delta thresholds to trigger background sync.
 - Reindex triggers: the index stores the embedding **provider/model + endpoint fingerprint + chunking params**. If any of those change, Clawdbot automatically resets and reindexes the entire store.

 ### Hybrid search (BM25 + vector)
@@ -299,11 +299,29 @@ agents: {

 Notes:
 - Session indexing is **opt-in** (off by default).
- Session updates are debounced and indexed lazily on the next `memory_search` (or manual `clawdbot memory index`).
+- Session updates are debounced and **indexed asynchronously** once they cross delta thresholds (best-effort).
+- `memory_search` never blocks on indexing; results can be slightly stale until background sync finishes.
 - Results still include snippets only; `memory_get` remains limited to memory files.
 - Session indexing is isolated per agent (only that agent’s session logs are indexed).
 - Session logs live on disk (`~/.clawdbot/agents/<agentId>/sessions/*.jsonl`). Any process/user with filesystem access can read them, so treat disk access as the trust boundary. For stricter isolation, run agents under separate OS users or hosts.

+Delta thresholds (defaults shown):
+
+```json5
+agents: {
+  defaults: {
+    memorySearch: {
+      sync: {
+        sessions: {
+          deltaBytes: 100000,   // ~100 KB
+          deltaMessages: 50     // JSONL lines
+        }
+      }
+    }
+  }
+}
+```
+
 ### SQLite vector acceleration (sqlite-vec)

 When the sqlite-vec extension is available, Clawdbot stores embeddings in a
@@ -24,7 +24,7 @@ Unknown keys, malformed types, or invalid values cause the Gateway to **refuse t

 When validation fails:
 - The Gateway does not boot.
- Only diagnostic commands are allowed (for example: `clawdbot doctor`, `clawdbot logs`, `clawdbot health`, `clawdbot status`, `clawdbot service`, `clawdbot help`).
+- Only diagnostic commands are allowed (for example: `clawdbot doctor`, `clawdbot logs`, `clawdbot health`, `clawdbot status`, `clawdbot daemon`, `clawdbot help`).
 - Run `clawdbot doctor` to see the exact issues.
 - Run `clawdbot doctor --fix` (or `--yes`) to apply migrations/repairs.

@@ -3022,6 +3022,9 @@ Template placeholders are expanded in `tools.media.*.models[].args` and `tools.m
 | `{{From}}` | Sender identifier (E.164 for WhatsApp; may differ per channel) |
 | `{{To}}` | Destination identifier |
 | `{{MessageSid}}` | Channel message id (when available) |
+| `{{MessageSidFull}}` | Provider-specific full message id when `MessageSid` is shortened |
+| `{{ReplyToId}}` | Reply-to message id (when available) |
+| `{{ReplyToIdFull}}` | Provider-specific full reply-to id when `ReplyToId` is shortened |
 | `{{SessionId}}` | Current session UUID |
 | `{{IsNewSession}}` | `"true"` when a new session was created |
 | `{{MediaUrl}}` | Inbound media pseudo-URL (if present) |
@@ -354,7 +354,7 @@ clawdbot doctor --fix
 Notes:
 - `clawdbot doctor` reports every invalid entry.
 - `clawdbot doctor --fix` applies migrations/repairs and rewrites the config.
- Diagnostic commands like `clawdbot logs`, `clawdbot health`, `clawdbot status`, and `clawdbot service` still run even if the config is invalid.
+- Diagnostic commands like `clawdbot logs`, `clawdbot health`, `clawdbot status`, and `clawdbot daemon` still run even if the config is invalid.

 ### “All models failed” — what should I check first?

@@ -50,14 +50,14 @@ forwards `exec` calls to the **node host** when `host=node` is selected.
 On the node machine:

 ```bash
-clawdbot node start --host <gateway-host> --port 18789 --display-name "Build Node"
+clawdbot node run --host <gateway-host> --port 18789 --display-name "Build Node"
 ```

 ### Start a node host (service)

 ```bash
-clawdbot node service install --host <gateway-host> --port 18789 --display-name "Build Node"
-clawdbot node service start
+clawdbot node install --host <gateway-host> --port 18789 --display-name "Build Node"
+clawdbot node start
 ```

 ### Pair + name
@@ -71,7 +71,7 @@ clawdbot nodes list
 ```

 Naming options:
- `--display-name` on `clawdbot node start/service install` (persists in `~/.clawdbot/node.json` on the node).
+- `--display-name` on `clawdbot node run` / `clawdbot node install` (persists in `~/.clawdbot/node.json` on the node).
 - `clawdbot nodes rename --node <id|name|ip> --name "Build Node"` (gateway override).

 ### Allowlist the commands
@@ -281,7 +281,7 @@ or for running a minimal node alongside a server.
 Start it:

 ```bash
-clawdbot node start --host <gateway-host> --port 18790
+clawdbot node run --host <gateway-host> --port 18790
 ```

 Notes:
@@ -5,7 +5,7 @@ read_when:
 ---
 # Clawdbot macOS IPC architecture

-**Current model:** a local Unix socket connects the **node service** to the **macOS app** for exec approvals + `system.run`. A `clawdbot-mac` debug CLI exists for discovery/connect checks; agent actions still flow through the Gateway WebSocket and `node.invoke`. UI automation uses PeekabooBridge.
+**Current model:** a local Unix socket connects the **node host service** to the **macOS app** for exec approvals + `system.run`. A `clawdbot-mac` debug CLI exists for discovery/connect checks; agent actions still flow through the Gateway WebSocket and `node.invoke`. UI automation uses PeekabooBridge.

 ## Goals
 - Single GUI app instance that owns all TCC-facing work (notifications, screen recording, mic, speech, AppleScript).
@@ -18,7 +18,7 @@ read_when:
 - Agent actions are performed via `node.invoke` (e.g. `system.run`, `system.notify`, `canvas.*`).

 ### Node service + app IPC
- A headless node service connects to the Gateway bridge.
+- A headless node host service connects to the Gateway bridge.
 - `system.run` requests are forwarded to the macOS app over a local Unix socket.
 - The app performs the exec in UI context, prompts if needed, and returns output.

@@ -57,7 +57,7 @@ The macOS app presents itself as a node. Common commands:
 The node reports a `permissions` map so agents can decide what’s allowed.

 Node service + app IPC:
- When the headless node service is running (remote mode), it connects to the Gateway WS as a node.
+- When the headless node host service is running (remote mode), it connects to the Gateway WS as a node.
 - `system.run` executes in the macOS app (UI/TCC context) over a local Unix socket; prompts + output stay in-app.

 Diagram (SCI):
@@ -30,7 +30,7 @@ read_when:
 - **Node identity:** use existing `nodeId`.
 - **Socket auth:** Unix socket + token (cross-platform); split later if needed.
 - **Node host state:** `~/.clawdbot/node.json` (node id + pairing token).
- **macOS exec host:** run `system.run` inside the macOS app; node service forwards requests over local IPC.
+- **macOS exec host:** run `system.run` inside the macOS app; node host service forwards requests over local IPC.
 - **No XPC helper:** stick to Unix socket + token + peer checks.

 ## Key concepts
@@ -54,7 +54,7 @@ Allowed (diagnostic-only):
 - `clawdbot health`
 - `clawdbot help`
 - `clawdbot status`
- `clawdbot service`
+- `clawdbot daemon`

 Everything else must hard-fail with: “Config invalid. Run `clawdbot doctor --fix`.”

@@ -26,6 +26,7 @@ When the operator says “release”, immediately do this preflight (no extra qu
 2) **Build & artifacts**
 - [ ] If A2UI inputs changed, run `pnpm canvas:a2ui:bundle` and commit any updated [`src/canvas-host/a2ui/a2ui.bundle.js`](https://github.com/clawdbot/clawdbot/blob/main/src/canvas-host/a2ui/a2ui.bundle.js).
 - [ ] `pnpm run build` (regenerates `dist/`).
+- [ ] Verify npm package `files` includes all required `dist/*` folders (notably `dist/node-host/**` and `dist/acp/**` for headless node + ACP CLI).
 - [ ] Confirm `dist/build-info.json` exists and includes the expected `commit` hash (CLI banner uses this for npm installs).
 - [ ] Optional: `npm pack --pack-destination /tmp` after the build; inspect the tarball contents and keep it handy for the GitHub release (do **not** commit it).

@@ -23,7 +23,7 @@ Exec approvals are enforced locally on the execution host:
 - **node host** → node runner (macOS companion app or headless node host)

 Planned macOS split:
- **node service** forwards `system.run` to the **macOS app** over local IPC.
+- **node host service** forwards `system.run` to the **macOS app** over local IPC.
 - **macOS app** enforces approvals + executes the command in UI context.

 ## Settings and storage
@@ -181,7 +181,7 @@ Notes:
 - If `process` is disallowed, `exec` runs synchronously and ignores `yieldMs`/`background`.
 - `elevated` is gated by `tools.elevated` plus any `agents.list[].tools.elevated` override (both must allow) and is an alias for `host=gateway` + `security=full`.
 - `elevated` only changes behavior when the agent is sandboxed (otherwise it’s a no-op).
- `host=node` can target a macOS companion app or a headless node host (`clawdbot node start`).
+- `host=node` can target a macOS companion app or a headless node host (`clawdbot node run`).
 - gateway/node approvals and allowlists: [Exec approvals](/tools/exec-approvals).

 ### `process`
@@ -1,10 +1,12 @@
 {
  "name": "@clawdbot/bluebubbles",
-  "version": "2026.1.15",
+  "version": "2026.1.21-1",
  "type": "module",
  "description": "Clawdbot BlueBubbles channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"],
+    "extensions": [
+      "./index.ts"
+    ],
    "channel": {
      "id": "bluebubbles",
      "label": "BlueBubbles",
@@ -38,6 +38,10 @@ vi.mock("./attachments.js", () => ({
  sendBlueBubblesAttachment: vi.fn().mockResolvedValue({ messageId: "att-msg-123" }),
 }));

+vi.mock("./monitor.js", () => ({
+  resolveBlueBubblesMessageId: vi.fn((id: string) => id),
+}));
+
 describe("bluebubblesMessageActions", () => {
  beforeEach(() => {
    vi.clearAllMocks();
@@ -358,6 +362,106 @@ describe("bluebubblesMessageActions", () => {
      );
    });

+    it("uses toolContext currentChannelId when no explicit target is provided", async () => {
+      const { sendBlueBubblesReaction } = await import("./reactions.js");
+      const { resolveChatGuidForTarget } = await import("./send.js");
+      vi.mocked(resolveChatGuidForTarget).mockResolvedValueOnce("iMessage;-;+15550001111");
+
+      const cfg: ClawdbotConfig = {
+        channels: {
+          bluebubbles: {
+            serverUrl: "http://localhost:1234",
+            password: "test-password",
+          },
+        },
+      };
+      await bluebubblesMessageActions.handleAction({
+        action: "react",
+        params: {
+          emoji: "👍",
+          messageId: "msg-456",
+        },
+        cfg,
+        accountId: null,
+        toolContext: {
+          currentChannelId: "bluebubbles:chat_guid:iMessage;-;+15550001111",
+        },
+      });
+
+      expect(resolveChatGuidForTarget).toHaveBeenCalledWith(
+        expect.objectContaining({
+          target: { kind: "chat_guid", chatGuid: "iMessage;-;+15550001111" },
+        }),
+      );
+      expect(sendBlueBubblesReaction).toHaveBeenCalledWith(
+        expect.objectContaining({
+          chatGuid: "iMessage;-;+15550001111",
+        }),
+      );
+    });
+
+    it("resolves short messageId before reacting", async () => {
+      const { resolveBlueBubblesMessageId } = await import("./monitor.js");
+      const { sendBlueBubblesReaction } = await import("./reactions.js");
+      vi.mocked(resolveBlueBubblesMessageId).mockReturnValueOnce("resolved-uuid");
+
+      const cfg: ClawdbotConfig = {
+        channels: {
+          bluebubbles: {
+            serverUrl: "http://localhost:1234",
+            password: "test-password",
+          },
+        },
+      };
+
+      await bluebubblesMessageActions.handleAction({
+        action: "react",
+        params: {
+          emoji: "❤️",
+          messageId: "1",
+          chatGuid: "iMessage;-;+15551234567",
+        },
+        cfg,
+        accountId: null,
+      });
+
+      expect(resolveBlueBubblesMessageId).toHaveBeenCalledWith("1", { requireKnownShortId: true });
+      expect(sendBlueBubblesReaction).toHaveBeenCalledWith(
+        expect.objectContaining({
+          messageGuid: "resolved-uuid",
+        }),
+      );
+    });
+
+    it("propagates short-id errors from the resolver", async () => {
+      const { resolveBlueBubblesMessageId } = await import("./monitor.js");
+      vi.mocked(resolveBlueBubblesMessageId).mockImplementationOnce(() => {
+        throw new Error("short id expired");
+      });
+
+      const cfg: ClawdbotConfig = {
+        channels: {
+          bluebubbles: {
+            serverUrl: "http://localhost:1234",
+            password: "test-password",
+          },
+        },
+      };
+
+      await expect(
+        bluebubblesMessageActions.handleAction({
+          action: "react",
+          params: {
+            emoji: "❤️",
+            messageId: "999",
+            chatGuid: "iMessage;-;+15551234567",
+          },
+          cfg,
+          accountId: null,
+        }),
+      ).rejects.toThrow("short id expired");
+    });
+
    it("accepts message param for edit action", async () => {
      const { editBlueBubblesMessage } = await import("./chat.js");

@@ -14,6 +14,7 @@ import {
 } from "clawdbot/plugin-sdk";

 import { resolveBlueBubblesAccount } from "./accounts.js";
+import { resolveBlueBubblesMessageId } from "./monitor.js";
 import { isMacOS26OrHigher } from "./probe.js";
 import { sendBlueBubblesReaction } from "./reactions.js";
 import { resolveChatGuidForTarget, sendMessageBlueBubbles } from "./send.js";
@@ -77,7 +78,7 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
    const accountId = typeof args.accountId === "string" ? args.accountId.trim() : undefined;
    return { to, accountId };
  },
-  handleAction: async ({ action, params, cfg, accountId }) => {
+  handleAction: async ({ action, params, cfg, accountId, toolContext }) => {
    const account = resolveBlueBubblesAccount({
      cfg: cfg as ClawdbotConfig,
      accountId: accountId ?? undefined,
@@ -86,7 +87,7 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
    const password = account.config.password?.trim();
    const opts = { cfg: cfg as ClawdbotConfig, accountId: accountId ?? undefined };

-    // Helper to resolve chatGuid from various params
+    // Helper to resolve chatGuid from various params or session context
    const resolveChatGuid = async (): Promise<string> => {
      const chatGuid = readStringParam(params, "chatGuid");
      if (chatGuid?.trim()) return chatGuid.trim();
@@ -94,6 +95,8 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
      const chatIdentifier = readStringParam(params, "chatIdentifier");
      const chatId = readNumberParam(params, "chatId", { integer: true });
      const to = readStringParam(params, "to");
+      // Fall back to session context if no explicit target provided
+      const contextTarget = toolContext?.currentChannelId?.trim();

      const target = chatIdentifier?.trim()
        ? ({
@@ -104,7 +107,9 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
          ? ({ kind: "chat_id", chatId } as BlueBubblesSendTarget)
          : to
            ? mapTarget(to)
-            : null;
+            : contextTarget
+              ? mapTarget(contextTarget)
+              : null;

      if (!target) {
        throw new Error(`BlueBubbles ${action} requires chatGuid, chatIdentifier, chatId, or to.`);
@@ -127,16 +132,18 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
      });
      if (isEmpty && !remove) {
        throw new Error(
-          "BlueBubbles react requires emoji parameter. Use action=react with emoji=<emoji> and messageId=<message_guid>.",
+          "BlueBubbles react requires emoji parameter. Use action=react with emoji=<emoji> and messageId=<message_id>.",
        );
      }
-      const messageId = readStringParam(params, "messageId");
-      if (!messageId) {
+      const rawMessageId = readStringParam(params, "messageId");
+      if (!rawMessageId) {
        throw new Error(
-          "BlueBubbles react requires messageId parameter (the message GUID to react to). " +
-            "Use action=react with messageId=<message_guid>, emoji=<emoji>, and to/chatGuid to identify the chat.",
+          "BlueBubbles react requires messageId parameter (the message ID to react to). " +
+            "Use action=react with messageId=<message_id>, emoji=<emoji>, and to/chatGuid to identify the chat.",
        );
      }
+      // Resolve short ID (e.g., "1", "2") to full UUID
+      const messageId = resolveBlueBubblesMessageId(rawMessageId, { requireKnownShortId: true });
      const partIndex = readNumberParam(params, "partIndex", { integer: true });
      const resolvedChatGuid = await resolveChatGuid();

@@ -161,20 +168,22 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
            "Apple removed the ability to edit iMessages in this version.",
        );
      }
-      const messageId = readStringParam(params, "messageId");
+      const rawMessageId = readStringParam(params, "messageId");
      const newText =
        readStringParam(params, "text") ??
        readStringParam(params, "newText") ??
        readStringParam(params, "message");
-      if (!messageId || !newText) {
+      if (!rawMessageId || !newText) {
        const missing: string[] = [];
-        if (!messageId) missing.push("messageId (the message GUID to edit)");
+        if (!rawMessageId) missing.push("messageId (the message ID to edit)");
        if (!newText) missing.push("text (the new message content)");
        throw new Error(
          `BlueBubbles edit requires: ${missing.join(", ")}. ` +
-            `Use action=edit with messageId=<message_guid>, text=<new_content>.`,
+            `Use action=edit with messageId=<message_id>, text=<new_content>.`,
        );
      }
+      // Resolve short ID (e.g., "1", "2") to full UUID
+      const messageId = resolveBlueBubblesMessageId(rawMessageId, { requireKnownShortId: true });
      const partIndex = readNumberParam(params, "partIndex", { integer: true });
      const backwardsCompatMessage = readStringParam(params, "backwardsCompatMessage");

@@ -184,18 +193,20 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
        backwardsCompatMessage: backwardsCompatMessage ?? undefined,
      });

-      return jsonResult({ ok: true, edited: messageId });
+      return jsonResult({ ok: true, edited: rawMessageId });
    }

    // Handle unsend action
    if (action === "unsend") {
-      const messageId = readStringParam(params, "messageId");
-      if (!messageId) {
+      const rawMessageId = readStringParam(params, "messageId");
+      if (!rawMessageId) {
        throw new Error(
-          "BlueBubbles unsend requires messageId parameter (the message GUID to unsend). " +
-            "Use action=unsend with messageId=<message_guid>.",
+          "BlueBubbles unsend requires messageId parameter (the message ID to unsend). " +
+            "Use action=unsend with messageId=<message_id>.",
        );
      }
+      // Resolve short ID (e.g., "1", "2") to full UUID
+      const messageId = resolveBlueBubblesMessageId(rawMessageId, { requireKnownShortId: true });
      const partIndex = readNumberParam(params, "partIndex", { integer: true });

      await unsendBlueBubblesMessage(messageId, {
@@ -203,24 +214,26 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
        partIndex: typeof partIndex === "number" ? partIndex : undefined,
      });

-      return jsonResult({ ok: true, unsent: messageId });
+      return jsonResult({ ok: true, unsent: rawMessageId });
    }

    // Handle reply action
    if (action === "reply") {
-      const messageId = readStringParam(params, "messageId");
+      const rawMessageId = readStringParam(params, "messageId");
      const text = readMessageText(params);
      const to = readStringParam(params, "to") ?? readStringParam(params, "target");
-      if (!messageId || !text || !to) {
+      if (!rawMessageId || !text || !to) {
        const missing: string[] = [];
-        if (!messageId) missing.push("messageId (the message GUID to reply to)");
+        if (!rawMessageId) missing.push("messageId (the message ID to reply to)");
        if (!text) missing.push("text or message (the reply message content)");
        if (!to) missing.push("to or target (the chat target)");
        throw new Error(
          `BlueBubbles reply requires: ${missing.join(", ")}. ` +
-            `Use action=reply with messageId=<message_guid>, message=<your reply>, target=<chat_target>.`,
+            `Use action=reply with messageId=<message_id>, message=<your reply>, target=<chat_target>.`,
        );
      }
+      // Resolve short ID (e.g., "1", "2") to full UUID
+      const messageId = resolveBlueBubblesMessageId(rawMessageId, { requireKnownShortId: true });
      const partIndex = readNumberParam(params, "partIndex", { integer: true });

      const result = await sendMessageBlueBubbles(to, text, {
@@ -229,7 +242,7 @@ export const bluebubblesMessageActions: ChannelMessageActionAdapter = {
        replyToPartIndex: typeof partIndex === "number" ? partIndex : undefined,
      });

-      return jsonResult({ ok: true, messageId: result.messageId, repliedTo: messageId });
+      return jsonResult({ ok: true, messageId: result.messageId, repliedTo: rawMessageId });
    }

    // Handle sendWithEffect action
@@ -20,6 +20,7 @@ import {
  resolveDefaultBlueBubblesAccountId,
 } from "./accounts.js";
 import { BlueBubblesConfigSchema } from "./config-schema.js";
+import { resolveBlueBubblesMessageId } from "./monitor.js";
 import { probeBlueBubbles, type BlueBubblesProbe } from "./probe.js";
 import { sendMessageBlueBubbles } from "./send.js";
 import {
@@ -65,7 +66,7 @@ export const bluebubblesPlugin: ChannelPlugin<ResolvedBlueBubblesAccount> = {
  threading: {
    buildToolContext: ({ context, hasRepliedRef }) => ({
      currentChannelId: context.To?.trim() || undefined,
-      currentThreadTs: context.ReplyToId,
+      currentThreadTs: context.ReplyToIdFull ?? context.ReplyToId,
      hasRepliedRef,
    }),
  },
@@ -237,7 +238,11 @@ export const bluebubblesPlugin: ChannelPlugin<ResolvedBlueBubblesAccount> = {
      return { ok: true, to: trimmed };
    },
    sendText: async ({ cfg, to, text, accountId, replyToId }) => {
-      const replyToMessageGuid = typeof replyToId === "string" ? replyToId.trim() : "";
+      const rawReplyToId = typeof replyToId === "string" ? replyToId.trim() : "";
+      // Resolve short ID (e.g., "5") to full UUID
+      const replyToMessageGuid = rawReplyToId
+        ? resolveBlueBubblesMessageId(rawReplyToId, { requireKnownShortId: true })
+        : "";
      const result = await sendMessageBlueBubbles(to, text, {
        cfg: cfg as ClawdbotConfig,
        accountId: accountId ?? undefined,
@@ -4,8 +4,9 @@ import { fileURLToPath } from "node:url";
 import { resolveChannelMediaMaxBytes, type ClawdbotConfig } from "clawdbot/plugin-sdk";

 import { sendBlueBubblesAttachment } from "./attachments.js";
-import { sendMessageBlueBubbles } from "./send.js";
+import { resolveBlueBubblesMessageId } from "./monitor.js";
 import { getBlueBubblesRuntime } from "./runtime.js";
+import { sendMessageBlueBubbles } from "./send.js";

 const HTTP_URL_RE = /^https?:\/\//i;
 const MB = 1024 * 1024;
@@ -134,12 +135,17 @@ export async function sendBlueBubblesMedia(params: {
    }
  }

+  // Resolve short ID (e.g., "5") to full UUID
+  const replyToMessageGuid = replyToId?.trim()
+    ? resolveBlueBubblesMessageId(replyToId.trim(), { requireKnownShortId: true })
+    : undefined;
+
  const attachmentResult = await sendBlueBubblesAttachment({
    to,
    buffer,
    filename: resolvedFilename ?? "attachment",
    contentType: resolvedContentType ?? undefined,
-    replyToMessageGuid: replyToId?.trim() || undefined,
+    replyToMessageGuid,
    opts: {
      cfg,
      accountId,
@@ -151,7 +157,7 @@ export async function sendBlueBubblesMedia(params: {
    await sendMessageBlueBubbles(to, trimmedCaption, {
      cfg,
      accountId,
-      replyToMessageGuid: replyToId?.trim() || undefined,
+      replyToMessageGuid,
    });
  }

@@ -6,6 +6,8 @@ import type { ClawdbotConfig, PluginRuntime } from "clawdbot/plugin-sdk";
 import {
  handleBlueBubblesWebhookRequest,
  registerBlueBubblesWebhookTarget,
+  resolveBlueBubblesMessageId,
+  _resetBlueBubblesShortIdState,
 } from "./monitor.js";
 import { setBlueBubblesRuntime } from "./runtime.js";
 import type { ResolvedBlueBubblesAccount } from "./accounts.js";
@@ -223,6 +225,8 @@ describe("BlueBubbles webhook monitor", () => {

  beforeEach(() => {
    vi.clearAllMocks();
+    // Reset short ID state between tests for predictable behavior
+    _resetBlueBubblesShortIdState();
    mockReadAllowFromStore.mockResolvedValue([]);
    mockUpsertPairingRequest.mockResolvedValue({ code: "TESTCODE", created: true });
    mockResolveRequireMention.mockReturnValue(false);
@@ -467,6 +471,98 @@ describe("BlueBubbles webhook monitor", () => {

      expect(handled).toBe(false);
    });
+
+    it("parses chatId when provided as a string (webhook variant)", async () => {
+      const { resolveChatGuidForTarget } = await import("./send.js");
+      vi.mocked(resolveChatGuidForTarget).mockClear();
+
+      const account = createMockAccount({ groupPolicy: "open" });
+      const config: ClawdbotConfig = {};
+      const core = createMockRuntime();
+      setBlueBubblesRuntime(core);
+
+      unregister = registerBlueBubblesWebhookTarget({
+        account,
+        config,
+        runtime: { log: vi.fn(), error: vi.fn() },
+        core,
+        path: "/bluebubbles-webhook",
+      });
+
+      const payload = {
+        type: "new-message",
+        data: {
+          text: "hello from group",
+          handle: { address: "+15551234567" },
+          isGroup: true,
+          isFromMe: false,
+          guid: "msg-1",
+          chatId: "123",
+          date: Date.now(),
+        },
+      };
+
+      const req = createMockRequest("POST", "/bluebubbles-webhook", payload);
+      const res = createMockResponse();
+
+      await handleBlueBubblesWebhookRequest(req, res);
+      await new Promise((resolve) => setTimeout(resolve, 50));
+
+      expect(resolveChatGuidForTarget).toHaveBeenCalledWith(
+        expect.objectContaining({
+          target: { kind: "chat_id", chatId: 123 },
+        }),
+      );
+    });
+
+    it("extracts chatGuid from nested chat object fields (webhook variant)", async () => {
+      const { sendMessageBlueBubbles, resolveChatGuidForTarget } = await import("./send.js");
+      vi.mocked(sendMessageBlueBubbles).mockClear();
+      vi.mocked(resolveChatGuidForTarget).mockClear();
+
+      mockDispatchReplyWithBufferedBlockDispatcher.mockImplementationOnce(async (params) => {
+        await params.dispatcherOptions.deliver({ text: "replying now" }, { kind: "final" });
+      });
+
+      const account = createMockAccount({ groupPolicy: "open" });
+      const config: ClawdbotConfig = {};
+      const core = createMockRuntime();
+      setBlueBubblesRuntime(core);
+
+      unregister = registerBlueBubblesWebhookTarget({
+        account,
+        config,
+        runtime: { log: vi.fn(), error: vi.fn() },
+        core,
+        path: "/bluebubbles-webhook",
+      });
+
+      const payload = {
+        type: "new-message",
+        data: {
+          text: "hello from group",
+          handle: { address: "+15551234567" },
+          isGroup: true,
+          isFromMe: false,
+          guid: "msg-1",
+          chat: { chatGuid: "iMessage;+;chat123456" },
+          date: Date.now(),
+        },
+      };
+
+      const req = createMockRequest("POST", "/bluebubbles-webhook", payload);
+      const res = createMockResponse();
+
+      await handleBlueBubblesWebhookRequest(req, res);
+      await new Promise((resolve) => setTimeout(resolve, 50));
+
+      expect(resolveChatGuidForTarget).not.toHaveBeenCalled();
+      expect(sendMessageBlueBubbles).toHaveBeenCalledWith(
+        "chat_guid:iMessage;+;chat123456",
+        expect.any(String),
+        expect.any(Object),
+      );
+    });
  });

  describe("DM pairing behavior vs allowFrom", () => {
@@ -1075,13 +1171,86 @@ describe("BlueBubbles webhook monitor", () => {

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
+      // ReplyToId is the full UUID since it wasn't previously cached
      expect(callArgs.ctx.ReplyToId).toBe("msg-0");
      expect(callArgs.ctx.ReplyToBody).toBe("original message");
      expect(callArgs.ctx.ReplyToSender).toBe("+15550000000");
-      expect(callArgs.ctx.Body).toContain("[Replying to +15550000000 id:msg-0]");
+      // Body uses just the ID (no sender) for token savings
+      expect(callArgs.ctx.Body).toContain("[Replying to id:msg-0]");
      expect(callArgs.ctx.Body).toContain("original message");
    });

+    it("hydrates missing reply sender/body from the recent-message cache", async () => {
+      const account = createMockAccount({ dmPolicy: "open", groupPolicy: "open" });
+      const config: ClawdbotConfig = {};
+      const core = createMockRuntime();
+      setBlueBubblesRuntime(core);
+
+      unregister = registerBlueBubblesWebhookTarget({
+        account,
+        config,
+        runtime: { log: vi.fn(), error: vi.fn() },
+        core,
+        path: "/bluebubbles-webhook",
+      });
+
+      const chatGuid = "iMessage;+;chat-reply-cache";
+
+      const originalPayload = {
+        type: "new-message",
+        data: {
+          text: "original message (cached)",
+          handle: { address: "+15550000000" },
+          isGroup: true,
+          isFromMe: false,
+          guid: "cache-msg-0",
+          chatGuid,
+          date: Date.now(),
+        },
+      };
+
+      const originalReq = createMockRequest("POST", "/bluebubbles-webhook", originalPayload);
+      const originalRes = createMockResponse();
+
+      await handleBlueBubblesWebhookRequest(originalReq, originalRes);
+      await new Promise((resolve) => setTimeout(resolve, 50));
+
+      // Only assert the reply message behavior below.
+      mockDispatchReplyWithBufferedBlockDispatcher.mockClear();
+
+      const replyPayload = {
+        type: "new-message",
+        data: {
+          text: "replying now",
+          handle: { address: "+15551234567" },
+          isGroup: true,
+          isFromMe: false,
+          guid: "cache-msg-1",
+          chatGuid,
+          // Only the GUID is provided; sender/body must be hydrated.
+          replyToMessageGuid: "cache-msg-0",
+          date: Date.now(),
+        },
+      };
+
+      const replyReq = createMockRequest("POST", "/bluebubbles-webhook", replyPayload);
+      const replyRes = createMockResponse();
+
+      await handleBlueBubblesWebhookRequest(replyReq, replyRes);
+      await new Promise((resolve) => setTimeout(resolve, 50));
+
+      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
+      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
+      // ReplyToId uses short ID "1" (first cached message) for token savings
+      expect(callArgs.ctx.ReplyToId).toBe("1");
+      expect(callArgs.ctx.ReplyToIdFull).toBe("cache-msg-0");
+      expect(callArgs.ctx.ReplyToBody).toBe("original message (cached)");
+      expect(callArgs.ctx.ReplyToSender).toBe("+15550000000");
+      // Body uses just the short ID (no sender) for token savings
+      expect(callArgs.ctx.Body).toContain("[Replying to id:1]");
+      expect(callArgs.ctx.Body).toContain("original message (cached)");
+    });
+
    it("falls back to threadOriginatorGuid when reply metadata is absent", async () => {
      const account = createMockAccount({ dmPolicy: "open" });
      const config: ClawdbotConfig = {};
@@ -1436,8 +1605,9 @@ describe("BlueBubbles webhook monitor", () => {
      await handleBlueBubblesWebhookRequest(req, res);
      await new Promise((resolve) => setTimeout(resolve, 50));

+      // Outbound message ID uses short ID "2" (inbound msg-1 is "1", outbound msg-123 is "2")
      expect(mockEnqueueSystemEvent).toHaveBeenCalledWith(
-        "BlueBubbles sent message id: msg-123",
+        'Assistant sent "replying now" [message_id:2]',
        expect.objectContaining({
          sessionKey: "agent:main:bluebubbles:dm:+15551234567",
        }),
@@ -1605,6 +1775,99 @@ describe("BlueBubbles webhook monitor", () => {
    });
  });

+  describe("short message ID mapping", () => {
+    it("assigns sequential short IDs to messages", async () => {
+      const account = createMockAccount({ dmPolicy: "open" });
+      const config: ClawdbotConfig = {};
+      const core = createMockRuntime();
+      setBlueBubblesRuntime(core);
+
+      unregister = registerBlueBubblesWebhookTarget({
+        account,
+        config,
+        runtime: { log: vi.fn(), error: vi.fn() },
+        core,
+        path: "/bluebubbles-webhook",
+      });
+
+      const payload = {
+        type: "new-message",
+        data: {
+          text: "hello",
+          handle: { address: "+15551234567" },
+          isGroup: false,
+          isFromMe: false,
+          guid: "msg-uuid-12345",
+          chatGuid: "iMessage;-;+15551234567",
+          date: Date.now(),
+        },
+      };
+
+      const req = createMockRequest("POST", "/bluebubbles-webhook", payload);
+      const res = createMockResponse();
+
+      await handleBlueBubblesWebhookRequest(req, res);
+      await new Promise((resolve) => setTimeout(resolve, 50));
+
+      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
+      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
+      // MessageSid should be short ID "1" instead of full UUID
+      expect(callArgs.ctx.MessageSid).toBe("1");
+      expect(callArgs.ctx.MessageSidFull).toBe("msg-uuid-12345");
+    });
+
+    it("resolves short ID back to UUID", async () => {
+      const account = createMockAccount({ dmPolicy: "open" });
+      const config: ClawdbotConfig = {};
+      const core = createMockRuntime();
+      setBlueBubblesRuntime(core);
+
+      unregister = registerBlueBubblesWebhookTarget({
+        account,
+        config,
+        runtime: { log: vi.fn(), error: vi.fn() },
+        core,
+        path: "/bluebubbles-webhook",
+      });
+
+      const payload = {
+        type: "new-message",
+        data: {
+          text: "hello",
+          handle: { address: "+15551234567" },
+          isGroup: false,
+          isFromMe: false,
+          guid: "msg-uuid-12345",
+          chatGuid: "iMessage;-;+15551234567",
+          date: Date.now(),
+        },
+      };
+
+      const req = createMockRequest("POST", "/bluebubbles-webhook", payload);
+      const res = createMockResponse();
+
+      await handleBlueBubblesWebhookRequest(req, res);
+      await new Promise((resolve) => setTimeout(resolve, 50));
+
+      // The short ID "1" should resolve back to the full UUID
+      expect(resolveBlueBubblesMessageId("1")).toBe("msg-uuid-12345");
+    });
+
+    it("returns UUID unchanged when not in cache", () => {
+      expect(resolveBlueBubblesMessageId("msg-not-cached")).toBe("msg-not-cached");
+    });
+
+    it("returns short ID unchanged when numeric but not in cache", () => {
+      expect(resolveBlueBubblesMessageId("999")).toBe("999");
+    });
+
+    it("throws when numeric short ID is missing and requireKnownShortId is set", () => {
+      expect(() =>
+        resolveBlueBubblesMessageId("999", { requireKnownShortId: true }),
+      ).toThrow(/short message id/i);
+    });
+  });
+
  describe("fromMe messages", () => {
    it("ignores messages from self (fromMe=true)", async () => {
      const account = createMockAccount();
@@ -1,11 +1,11 @@
 import type { IncomingMessage, ServerResponse } from "node:http";

 import type { ClawdbotConfig } from "clawdbot/plugin-sdk";
+import { resolveAckReaction } from "clawdbot/plugin-sdk";
 import { markBlueBubblesChatRead, sendBlueBubblesTyping } from "./chat.js";
 import { resolveChatGuidForTarget, sendMessageBlueBubbles } from "./send.js";
 import { downloadBlueBubblesAttachment } from "./attachments.js";
 import { formatBlueBubblesChatTarget, isAllowedBlueBubblesSender, normalizeBlueBubblesHandle } from "./targets.js";
-import { resolveAckReaction } from "../../../src/agents/identity.js";
 import { sendBlueBubblesMedia } from "./media-send.js";
 import type { BlueBubblesAccountConfig, BlueBubblesAttachment } from "./types.js";
 import type { ResolvedBlueBubblesAccount } from "./accounts.js";
@@ -31,6 +31,173 @@ const DEFAULT_WEBHOOK_PATH = "/bluebubbles-webhook";
 const DEFAULT_TEXT_LIMIT = 4000;
 const invalidAckReactions = new Set<string>();

+const REPLY_CACHE_MAX = 2000;
+const REPLY_CACHE_TTL_MS = 6 * 60 * 60 * 1000;
+
+type BlueBubblesReplyCacheEntry = {
+  accountId: string;
+  messageId: string;
+  shortId: string;
+  chatGuid?: string;
+  chatIdentifier?: string;
+  chatId?: number;
+  senderLabel?: string;
+  body?: string;
+  timestamp: number;
+};
+
+// Best-effort cache for resolving reply context when BlueBubbles webhooks omit sender/body.
+const blueBubblesReplyCacheByMessageId = new Map<string, BlueBubblesReplyCacheEntry>();
+
+// Bidirectional maps for short ID ↔ UUID resolution (token savings optimization)
+const blueBubblesShortIdToUuid = new Map<string, string>();
+const blueBubblesUuidToShortId = new Map<string, string>();
+let blueBubblesShortIdCounter = 0;
+
+function trimOrUndefined(value?: string | null): string | undefined {
+  const trimmed = value?.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function generateShortId(): string {
+  blueBubblesShortIdCounter += 1;
+  return String(blueBubblesShortIdCounter);
+}
+
+function rememberBlueBubblesReplyCache(
+  entry: Omit<BlueBubblesReplyCacheEntry, "shortId">,
+): BlueBubblesReplyCacheEntry {
+  const messageId = entry.messageId.trim();
+  if (!messageId) {
+    return { ...entry, shortId: "" };
+  }
+
+  // Check if we already have a short ID for this UUID
+  let shortId = blueBubblesUuidToShortId.get(messageId);
+  if (!shortId) {
+    shortId = generateShortId();
+    blueBubblesShortIdToUuid.set(shortId, messageId);
+    blueBubblesUuidToShortId.set(messageId, shortId);
+  }
+
+  const fullEntry: BlueBubblesReplyCacheEntry = { ...entry, shortId };
+
+  // Refresh insertion order.
+  blueBubblesReplyCacheByMessageId.delete(messageId);
+  blueBubblesReplyCacheByMessageId.set(messageId, fullEntry);
+
+  // Opportunistic prune.
+  const cutoff = Date.now() - REPLY_CACHE_TTL_MS;
+  for (const [key, value] of blueBubblesReplyCacheByMessageId) {
+    if (value.timestamp < cutoff) {
+      blueBubblesReplyCacheByMessageId.delete(key);
+      // Clean up short ID mappings for expired entries
+      if (value.shortId) {
+        blueBubblesShortIdToUuid.delete(value.shortId);
+        blueBubblesUuidToShortId.delete(key);
+      }
+      continue;
+    }
+    break;
+  }
+  while (blueBubblesReplyCacheByMessageId.size > REPLY_CACHE_MAX) {
+    const oldest = blueBubblesReplyCacheByMessageId.keys().next().value as string | undefined;
+    if (!oldest) break;
+    const oldEntry = blueBubblesReplyCacheByMessageId.get(oldest);
+    blueBubblesReplyCacheByMessageId.delete(oldest);
+    // Clean up short ID mappings for evicted entries
+    if (oldEntry?.shortId) {
+      blueBubblesShortIdToUuid.delete(oldEntry.shortId);
+      blueBubblesUuidToShortId.delete(oldest);
+    }
+  }
+
+  return fullEntry;
+}
+
+/**
+ * Resolves a short message ID (e.g., "1", "2") to a full BlueBubbles UUID.
+ * Returns the input unchanged if it's already a UUID or not found in the mapping.
+ */
+export function resolveBlueBubblesMessageId(
+  shortOrUuid: string,
+  opts?: { requireKnownShortId?: boolean },
+): string {
+  const trimmed = shortOrUuid.trim();
+  if (!trimmed) return trimmed;
+
+  // If it looks like a short ID (numeric), try to resolve it
+  if (/^\d+$/.test(trimmed)) {
+    const uuid = blueBubblesShortIdToUuid.get(trimmed);
+    if (uuid) return uuid;
+    if (opts?.requireKnownShortId) {
+      throw new Error(
+        `BlueBubbles short message id "${trimmed}" is no longer available. Use MessageSidFull.`,
+      );
+    }
+  }
+
+  // Return as-is (either already a UUID or not found)
+  return trimmed;
+}
+
+/**
+ * Resets the short ID state. Only use in tests.
+ * @internal
+ */
+export function _resetBlueBubblesShortIdState(): void {
+  blueBubblesShortIdToUuid.clear();
+  blueBubblesUuidToShortId.clear();
+  blueBubblesReplyCacheByMessageId.clear();
+  blueBubblesShortIdCounter = 0;
+}
+
+/**
+ * Gets the short ID for a UUID, if one exists.
+ */
+function getShortIdForUuid(uuid: string): string | undefined {
+  return blueBubblesUuidToShortId.get(uuid.trim());
+}
+
+function resolveReplyContextFromCache(params: {
+  accountId: string;
+  replyToId: string;
+  chatGuid?: string;
+  chatIdentifier?: string;
+  chatId?: number;
+}): BlueBubblesReplyCacheEntry | null {
+  const replyToId = params.replyToId.trim();
+  if (!replyToId) return null;
+
+  const cached = blueBubblesReplyCacheByMessageId.get(replyToId);
+  if (!cached) return null;
+  if (cached.accountId !== params.accountId) return null;
+
+  const cutoff = Date.now() - REPLY_CACHE_TTL_MS;
+  if (cached.timestamp < cutoff) {
+    blueBubblesReplyCacheByMessageId.delete(replyToId);
+    return null;
+  }
+
+  const chatGuid = trimOrUndefined(params.chatGuid);
+  const chatIdentifier = trimOrUndefined(params.chatIdentifier);
+  const cachedChatGuid = trimOrUndefined(cached.chatGuid);
+  const cachedChatIdentifier = trimOrUndefined(cached.chatIdentifier);
+  const chatId = typeof params.chatId === "number" ? params.chatId : undefined;
+  const cachedChatId = typeof cached.chatId === "number" ? cached.chatId : undefined;
+
+  // Avoid cross-chat collisions if we have identifiers.
+  if (chatGuid && cachedChatGuid && chatGuid !== cachedChatGuid) return null;
+  if (!chatGuid && chatIdentifier && cachedChatIdentifier && chatIdentifier !== cachedChatIdentifier) {
+    return null;
+  }
+  if (!chatGuid && !chatIdentifier && chatId && cachedChatId && chatId !== cachedChatId) {
+    return null;
+  }
+
+  return cached;
+}
+
 type BlueBubblesCoreRuntime = ReturnType<typeof getBlueBubblesRuntime>;

 function logVerbose(core: BlueBubblesCoreRuntime, runtime: BlueBubblesRuntimeEnv, message: string): void {
@@ -217,19 +384,29 @@ function buildMessagePlaceholder(message: NormalizedWebhookMessage): string {
  return "";
 }

+const REPLY_BODY_TRUNCATE_LENGTH = 60;
+
 function formatReplyContext(message: {
  replyToId?: string;
+  replyToShortId?: string;
  replyToBody?: string;
  replyToSender?: string;
 }): string | null {
  if (!message.replyToId && !message.replyToBody && !message.replyToSender) return null;
-  const sender = message.replyToSender?.trim() || "unknown sender";
-  const idPart = message.replyToId ? ` id:${message.replyToId}` : "";
-  const body = message.replyToBody?.trim();
-  if (!body) {
-    return `[Replying to ${sender}${idPart}]\n[/Replying]`;
+  // Prefer short ID for token savings
+  const displayId = message.replyToShortId || message.replyToId;
+  // Only include sender if we don't have an ID (fallback)
+  const label = displayId ? `id:${displayId}` : (message.replyToSender?.trim() || "unknown");
+  const rawBody = message.replyToBody?.trim();
+  if (!rawBody) {
+    return `[Replying to ${label}]\n[/Replying]`;
  }
-  return `[Replying to ${sender}${idPart}]\n${body}\n[/Replying]`;
+  // Truncate long reply bodies for token savings
+  const body =
+    rawBody.length > REPLY_BODY_TRUNCATE_LENGTH
+      ? `${rawBody.slice(0, REPLY_BODY_TRUNCATE_LENGTH)}…`
+      : rawBody;
+  return `[Replying to ${label}]\n${body}\n[/Replying]`;
 }

 function readNumberLike(record: Record<string, unknown> | null, key: string): number | undefined {
@@ -404,6 +581,15 @@ function resolveGroupFlagFromChatGuid(chatGuid?: string | null): boolean | undef
  return undefined;
 }

+function extractChatIdentifierFromChatGuid(chatGuid?: string | null): string | undefined {
+  const guid = chatGuid?.trim();
+  if (!guid) return undefined;
+  const parts = guid.split(";");
+  if (parts.length < 3) return undefined;
+  const identifier = parts[2]?.trim();
+  return identifier || undefined;
+}
+
 function formatGroupAllowlistEntry(params: {
  chatGuid?: string;
  chatId?: number;
@@ -550,20 +736,31 @@ function normalizeWebhookMessage(payload: Record<string, unknown>): NormalizedWe
  const chatGuid =
    readString(message, "chatGuid") ??
    readString(message, "chat_guid") ??
+    readString(chat, "chatGuid") ??
+    readString(chat, "chat_guid") ??
    readString(chat, "guid") ??
+    readString(chatFromList, "chatGuid") ??
+    readString(chatFromList, "chat_guid") ??
    readString(chatFromList, "guid");
  const chatIdentifier =
    readString(message, "chatIdentifier") ??
    readString(message, "chat_identifier") ??
+    readString(chat, "chatIdentifier") ??
+    readString(chat, "chat_identifier") ??
    readString(chat, "identifier") ??
    readString(chatFromList, "chatIdentifier") ??
    readString(chatFromList, "chat_identifier") ??
-    readString(chatFromList, "identifier");
+    readString(chatFromList, "identifier") ??
+    extractChatIdentifierFromChatGuid(chatGuid);
  const chatId =
-    readNumber(message, "chatId") ??
-    readNumber(message, "chat_id") ??
-    readNumber(chat, "id") ??
-    readNumber(chatFromList, "id");
+    readNumberLike(message, "chatId") ??
+    readNumberLike(message, "chat_id") ??
+    readNumberLike(chat, "chatId") ??
+    readNumberLike(chat, "chat_id") ??
+    readNumberLike(chat, "id") ??
+    readNumberLike(chatFromList, "chatId") ??
+    readNumberLike(chatFromList, "chat_id") ??
+    readNumberLike(chatFromList, "id");
  const chatName =
    readString(message, "chatName") ??
    readString(chat, "displayName") ??
@@ -679,19 +876,30 @@ function normalizeWebhookReaction(payload: Record<string, unknown>): NormalizedW
  const chatGuid =
    readString(message, "chatGuid") ??
    readString(message, "chat_guid") ??
+    readString(chat, "chatGuid") ??
+    readString(chat, "chat_guid") ??
    readString(chat, "guid") ??
+    readString(chatFromList, "chatGuid") ??
+    readString(chatFromList, "chat_guid") ??
    readString(chatFromList, "guid");
  const chatIdentifier =
    readString(message, "chatIdentifier") ??
    readString(message, "chat_identifier") ??
+    readString(chat, "chatIdentifier") ??
+    readString(chat, "chat_identifier") ??
    readString(chat, "identifier") ??
    readString(chatFromList, "chatIdentifier") ??
    readString(chatFromList, "chat_identifier") ??
-    readString(chatFromList, "identifier");
+    readString(chatFromList, "identifier") ??
+    extractChatIdentifierFromChatGuid(chatGuid);
  const chatId =
    readNumberLike(message, "chatId") ??
    readNumberLike(message, "chat_id") ??
+    readNumberLike(chat, "chatId") ??
+    readNumberLike(chat, "chat_id") ??
    readNumberLike(chat, "id") ??
+    readNumberLike(chatFromList, "chatId") ??
+    readNumberLike(chatFromList, "chat_id") ??
    readNumberLike(chatFromList, "id");
  const chatName =
    readString(message, "chatName") ??
@@ -901,14 +1109,39 @@ async function processMessage(
  target: WebhookTarget,
 ): Promise<void> {
  const { account, config, runtime, core, statusSink } = target;
-  if (message.fromMe) return;
+
  const groupFlag = resolveGroupFlagFromChatGuid(message.chatGuid);
  const isGroup = typeof groupFlag === "boolean" ? groupFlag : message.isGroup;

  const text = message.text.trim();
  const attachments = message.attachments ?? [];
  const placeholder = buildMessagePlaceholder(message);
-  if (!text && !placeholder) {
+  const rawBody = text || placeholder;
+
+  const cacheMessageId = message.messageId?.trim();
+  let messageShortId: string | undefined;
+  const cacheInboundMessage = () => {
+    if (!cacheMessageId) return;
+    const cacheEntry = rememberBlueBubblesReplyCache({
+      accountId: account.accountId,
+      messageId: cacheMessageId,
+      chatGuid: message.chatGuid,
+      chatIdentifier: message.chatIdentifier,
+      chatId: message.chatId,
+      senderLabel: message.fromMe ? "me" : message.senderId,
+      body: rawBody,
+      timestamp: message.timestamp ?? Date.now(),
+    });
+    messageShortId = cacheEntry.shortId;
+  };
+
+  if (message.fromMe) {
+    // Cache from-me messages so reply context can resolve sender/body.
+    cacheInboundMessage();
+    return;
+  }
+
+  if (!rawBody) {
    logVerbose(core, runtime, `drop: empty text sender=${message.senderId}`);
    return;
  }
@@ -1148,6 +1381,10 @@ async function processMessage(
    return;
  }

+  // Cache allowed inbound messages so later replies can resolve sender/body without
+  // surfacing dropped content (allowlist/mention/command gating).
+  cacheInboundMessage();
+
  const baseUrl = account.config.serverUrl?.trim();
  const password = account.config.password?.trim();
  const maxBytes =
@@ -1199,12 +1436,42 @@ async function processMessage(
      }
    }
  }
-  const rawBody = text.trim() || placeholder;
-  const replyContext = formatReplyContext(message);
+  let replyToId = message.replyToId;
+  let replyToBody = message.replyToBody;
+  let replyToSender = message.replyToSender;
+  let replyToShortId: string | undefined;
+
+  if (replyToId && (!replyToBody || !replyToSender)) {
+    const cached = resolveReplyContextFromCache({
+      accountId: account.accountId,
+      replyToId,
+      chatGuid: message.chatGuid,
+      chatIdentifier: message.chatIdentifier,
+      chatId: message.chatId,
+    });
+    if (cached) {
+      if (!replyToBody && cached.body) replyToBody = cached.body;
+      if (!replyToSender && cached.senderLabel) replyToSender = cached.senderLabel;
+      replyToShortId = cached.shortId;
+      if (core.logging.shouldLogVerbose()) {
+        const preview = (cached.body ?? "").replace(/\s+/g, " ").slice(0, 120);
+        logVerbose(
+          core,
+          runtime,
+          `reply-context cache hit replyToId=${replyToId} sender=${replyToSender ?? ""} body="${preview}"`,
+        );
+      }
+    }
+  }
+
+  // If no cached short ID, try to get one from the UUID directly
+  if (replyToId && !replyToShortId) {
+    replyToShortId = getShortIdForUuid(replyToId);
+  }
+
+  const replyContext = formatReplyContext({ replyToId, replyToShortId, replyToBody, replyToSender });
  const baseBody = replyContext ? `${rawBody}\n\n${replyContext}` : rawBody;
-  const fromLabel = isGroup
-    ? `group:${peerId}`
-    : message.senderName || `user:${message.senderId}`;
+  const fromLabel = isGroup ? undefined : message.senderName || `user:${message.senderId}`;
  const groupSubject = isGroup ? message.chatName?.trim() || undefined : undefined;
  const groupMembers = isGroup
    ? formatGroupMembers({
@@ -1230,12 +1497,12 @@ async function processMessage(
  });
  let chatGuidForActions = chatGuid;
  if (!chatGuidForActions && baseUrl && password) {
-  const target =
+    const target =
      isGroup && (chatId || chatIdentifier)
        ? chatId
-          ? { kind: "chat_id", chatId }
-          : { kind: "chat_identifier", chatIdentifier: chatIdentifier ?? "" }
-        : { kind: "handle", address: message.senderId };
+          ? ({ kind: "chat_id", chatId } as const)
+          : ({ kind: "chat_identifier", chatIdentifier: chatIdentifier ?? "" } as const)
+        : ({ kind: "handle", address: message.senderId } as const);
    if (target.kind !== "chat_identifier" || target.chatIdentifier) {
      chatGuidForActions =
        (await resolveChatGuidForTarget({
@@ -1316,10 +1583,23 @@ async function processMessage(
      ? formatBlueBubblesChatTarget({ chatGuid: chatGuidForActions })
      : message.senderId;

-  const maybeEnqueueOutboundMessageId = (messageId?: string) => {
+  const maybeEnqueueOutboundMessageId = (messageId?: string, snippet?: string) => {
    const trimmed = messageId?.trim();
    if (!trimmed || trimmed === "ok" || trimmed === "unknown") return;
-    core.system.enqueueSystemEvent(`BlueBubbles sent message id: ${trimmed}`, {
+    // Cache outbound message to get short ID
+    const cacheEntry = rememberBlueBubblesReplyCache({
+      accountId: account.accountId,
+      messageId: trimmed,
+      chatGuid: chatGuidForActions ?? chatGuid,
+      chatIdentifier,
+      chatId,
+      senderLabel: "me",
+      body: snippet ?? "",
+      timestamp: Date.now(),
+    });
+    const displayId = cacheEntry.shortId || trimmed;
+    const preview = snippet ? ` "${snippet.slice(0, 12)}${snippet.length > 12 ? "…" : ""}"` : "";
+    core.system.enqueueSystemEvent(`Assistant sent${preview} [message_id:${displayId}]`, {
      sessionKey: route.sessionKey,
      contextKey: `bluebubbles:outbound:${outboundTarget}:${trimmed}`,
    });
@@ -1343,16 +1623,20 @@ async function processMessage(
    AccountId: route.accountId,
    ChatType: isGroup ? "group" : "direct",
    ConversationLabel: fromLabel,
-    ReplyToId: message.replyToId,
-    ReplyToBody: message.replyToBody,
-    ReplyToSender: message.replyToSender,
+    // Use short ID for token savings (agent can use this to reference the message)
+    ReplyToId: replyToShortId || replyToId,
+    ReplyToIdFull: replyToId,
+    ReplyToBody: replyToBody,
+    ReplyToSender: replyToSender,
    GroupSubject: groupSubject,
    GroupMembers: groupMembers,
    SenderName: message.senderName || undefined,
    SenderId: message.senderId,
    Provider: "bluebubbles",
    Surface: "bluebubbles",
-    MessageSid: message.messageId,
+    // Use short ID for token savings (agent can use this to reference the message)
+    MessageSid: messageShortId || message.messageId,
+    MessageSidFull: message.messageId,
    Timestamp: message.timestamp,
    OriginatingChannel: "bluebubbles",
    OriginatingTo: `bluebubbles:${outboundTarget}`,
@@ -1367,6 +1651,11 @@ async function processMessage(
      cfg: config,
      dispatcherOptions: {
        deliver: async (payload) => {
+          const rawReplyToId = typeof payload.replyToId === "string" ? payload.replyToId.trim() : "";
+          // Resolve short ID (e.g., "5") to full UUID
+          const replyToMessageGuid = rawReplyToId
+            ? resolveBlueBubblesMessageId(rawReplyToId, { requireKnownShortId: true })
+            : "";
          const mediaList = payload.mediaUrls?.length
            ? payload.mediaUrls
            : payload.mediaUrl
@@ -1382,10 +1671,11 @@ async function processMessage(
                to: outboundTarget,
                mediaUrl,
                caption: caption ?? undefined,
-                replyToId: payload.replyToId ?? null,
+                replyToId: replyToMessageGuid || null,
                accountId: account.accountId,
              });
-              maybeEnqueueOutboundMessageId(result.messageId);
+              const cachedBody = (caption ?? "").trim() || "<media:attachment>";
+              maybeEnqueueOutboundMessageId(result.messageId, cachedBody);
              sentMessage = true;
              statusSink?.({ lastOutboundAt: Date.now() });
            }
@@ -1400,14 +1690,12 @@ async function processMessage(
          if (!chunks.length && payload.text) chunks.push(payload.text);
          if (!chunks.length) return;
          for (const chunk of chunks) {
-            const replyToMessageGuid =
-              typeof payload.replyToId === "string" ? payload.replyToId.trim() : "";
            const result = await sendMessageBlueBubbles(outboundTarget, chunk, {
              cfg: config,
              accountId: account.accountId,
              replyToMessageGuid: replyToMessageGuid || undefined,
            });
-            maybeEnqueueOutboundMessageId(result.messageId);
+            maybeEnqueueOutboundMessageId(result.messageId, chunk);
            sentMessage = true;
            statusSink?.({ lastOutboundAt: Date.now() });
          }
@@ -1541,7 +1829,9 @@ async function processReaction(

  const senderLabel = reaction.senderName || reaction.senderId;
  const chatLabel = reaction.isGroup ? ` in group:${peerId}` : "";
-  const text = `BlueBubbles reaction ${reaction.action}: ${reaction.emoji} by ${senderLabel}${chatLabel} on msg ${reaction.messageId}`;
+  // Use short ID for token savings
+  const messageDisplayId = getShortIdForUuid(reaction.messageId) || reaction.messageId;
+  const text = `BlueBubbles reaction ${reaction.action}: ${reaction.emoji} by ${senderLabel}${chatLabel} on msg ${messageDisplayId}`;
  core.system.enqueueSystemEvent(text, {
    sessionKey: route.sessionKey,
    contextKey: `bluebubbles:reaction:${reaction.action}:${peerId}:${reaction.messageId}:${reaction.senderId}:${reaction.emoji}`,
@@ -1,11 +1,17 @@
-import type { ClawdbotConfig, DmPolicy, WizardPrompter } from "clawdbot/plugin-sdk";
-import { DEFAULT_ACCOUNT_ID, normalizeAccountId } from "clawdbot/plugin-sdk";
 import type {
  ChannelOnboardingAdapter,
  ChannelOnboardingDmPolicy,
-} from "../../../src/channels/plugins/onboarding-types.js";
-import { addWildcardAllowFrom, promptAccountId } from "../../../src/channels/plugins/onboarding/helpers.js";
-import { formatDocsLink } from "../../../src/terminal/links.js";
+  ClawdbotConfig,
+  DmPolicy,
+  WizardPrompter,
+} from "clawdbot/plugin-sdk";
+import {
+  DEFAULT_ACCOUNT_ID,
+  addWildcardAllowFrom,
+  formatDocsLink,
+  normalizeAccountId,
+  promptAccountId,
+} from "clawdbot/plugin-sdk";
 import {
  listBlueBubblesAccountIds,
  resolveBlueBubblesAccount,
@@ -20,32 +20,101 @@ const REACTION_TYPES = new Set([
 ]);

 const REACTION_ALIASES = new Map<string, string>([
+  // General
  ["heart", "love"],
+  ["love", "love"],
+  ["❤", "love"],
+  ["❤️", "love"],
+  ["red_heart", "love"],
  ["thumbs_up", "like"],
-  ["thumbs-down", "dislike"],
+  ["thumbsup", "like"],
+  ["thumbs-up", "like"],
+  ["thumbsup", "like"],
+  ["like", "like"],
+  ["thumb", "like"],
+  ["ok", "like"],
  ["thumbs_down", "dislike"],
+  ["thumbsdown", "dislike"],
+  ["thumbs-down", "dislike"],
+  ["dislike", "dislike"],
+  ["boo", "dislike"],
+  ["no", "dislike"],
+  // Laugh
  ["haha", "laugh"],
  ["lol", "laugh"],
+  ["lmao", "laugh"],
+  ["rofl", "laugh"],
+  ["😂", "laugh"],
+  ["🤣", "laugh"],
+  ["xd", "laugh"],
+  ["laugh", "laugh"],
+  // Emphasize / exclaim
  ["emphasis", "emphasize"],
+  ["emphasize", "emphasize"],
  ["exclaim", "emphasize"],
+  ["!!", "emphasize"],
+  ["‼", "emphasize"],
+  ["‼️", "emphasize"],
+  ["❗", "emphasize"],
+  ["important", "emphasize"],
+  ["bang", "emphasize"],
+  // Question
  ["question", "question"],
+  ["?", "question"],
+  ["❓", "question"],
+  ["❔", "question"],
+  ["ask", "question"],
+  // Apple/Messages names
+  ["loved", "love"],
+  ["liked", "like"],
+  ["disliked", "dislike"],
+  ["laughed", "laugh"],
+  ["emphasized", "emphasize"],
+  ["questioned", "question"],
+  // Colloquial / informal
+  ["fire", "love"],
+  ["🔥", "love"],
+  ["wow", "emphasize"],
+  ["!", "emphasize"],
+  // Edge: generic emoji name forms
+  ["heart_eyes", "love"],
+  ["smile", "laugh"],
+  ["smiley", "laugh"],
+  ["happy", "laugh"],
+  ["joy", "laugh"],
 ]);

 const REACTION_EMOJIS = new Map<string, string>([
+  // Love
  ["❤️", "love"],
  ["❤", "love"],
  ["♥️", "love"],
+  ["♥", "love"],
  ["😍", "love"],
+  ["💕", "love"],
+  // Like
  ["👍", "like"],
+  ["👌", "like"],
+  // Dislike
  ["👎", "dislike"],
+  ["🙅", "dislike"],
+  // Laugh
  ["😂", "laugh"],
  ["🤣", "laugh"],
  ["😆", "laugh"],
+  ["😁", "laugh"],
+  ["😹", "laugh"],
+  // Emphasize
  ["‼️", "emphasize"],
  ["‼", "emphasize"],
+  ["!!", "emphasize"],
  ["❗", "emphasize"],
+  ["❕", "emphasize"],
+  ["!", "emphasize"],
+  // Question
  ["❓", "question"],
  ["❔", "question"],
+  ["?", "question"],
 ]);

 function resolveAccount(params: BlueBubblesReactionOpts) {
@@ -96,6 +96,33 @@ describe("send", () => {
      expect(result).toBe("iMessage;-;chat123");
    });

+    it("matches chat_identifier against the 3rd component of chat GUID", async () => {
+      mockFetch.mockResolvedValueOnce({
+        ok: true,
+        json: () =>
+          Promise.resolve({
+            data: [
+              {
+                guid: "iMessage;+;chat660250192681427962",
+                participants: [],
+              },
+            ],
+          }),
+      });
+
+      const target: BlueBubblesSendTarget = {
+        kind: "chat_identifier",
+        chatIdentifier: "chat660250192681427962",
+      };
+      const result = await resolveChatGuidForTarget({
+        baseUrl: "http://localhost:1234",
+        password: "test",
+        target,
+      });
+
+      expect(result).toBe("iMessage;+;chat660250192681427962");
+    });
+
    it("resolves handle target by matching participant", async () => {
      mockFetch.mockResolvedValueOnce({
        ok: true,
@@ -133,6 +133,13 @@ function extractChatId(chat: BlueBubblesChatRecord): number | null {
  return null;
 }

+function extractChatIdentifierFromChatGuid(chatGuid: string): string | null {
+  const parts = chatGuid.split(";");
+  if (parts.length < 3) return null;
+  const identifier = parts[2]?.trim();
+  return identifier ? identifier : null;
+}
+
 function extractParticipantAddresses(chat: BlueBubblesChatRecord): string[] {
  const raw =
    (Array.isArray(chat.participants) ? chat.participants : null) ??
@@ -223,7 +230,16 @@ export async function resolveChatGuidForTarget(params: {
      }
      if (targetChatIdentifier) {
        const guid = extractChatGuid(chat);
-        if (guid && guid === targetChatIdentifier) return guid;
+        if (guid) {
+          // Back-compat: some callers might pass a full chat GUID.
+          if (guid === targetChatIdentifier) return guid;
+
+          // Primary match: BlueBubbles `chat_identifier:*` targets correspond to the
+          // third component of the chat GUID: `service;(+|-) ;identifier`.
+          const guidIdentifier = extractChatIdentifierFromChatGuid(guid);
+          if (guidIdentifier && guidIdentifier === targetChatIdentifier) return guid;
+        }
+
        const identifier =
          typeof chat.identifier === "string"
            ? chat.identifier
@@ -232,7 +248,7 @@ export async function resolveChatGuidForTarget(params: {
              : typeof chat.chat_identifier === "string"
                ? chat.chat_identifier
                : "";
-        if (identifier && identifier === targetChatIdentifier) return extractChatGuid(chat);
+        if (identifier && identifier === targetChatIdentifier) return guid ?? extractChatGuid(chat);
      }
      if (normalizedHandle) {
        const guid = extractChatGuid(chat);
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/copilot-proxy",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Copilot Proxy provider plugin",
  "clawdbot": {
@@ -1,10 +1,12 @@
 {
  "name": "@clawdbot/diagnostics-otel",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot diagnostics OpenTelemetry exporter",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  },
  "dependencies": {
    "@opentelemetry/api": "^1.9.0",
@@ -1,9 +1,11 @@
 {
  "name": "@clawdbot/discord",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Discord channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  }
 }
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/google-antigravity-auth",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Google Antigravity OAuth provider plugin",
  "clawdbot": {
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/google-gemini-cli-auth",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Gemini CLI OAuth provider plugin",
  "clawdbot": {
@@ -1,9 +1,11 @@
 {
  "name": "@clawdbot/imessage",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot iMessage channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  }
 }
@@ -1,5 +1,15 @@
 # Changelog

+## 2026.1.20-2
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
+## 2026.1.20
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.17-1

 ### Changes
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/matrix",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Matrix channel plugin",
  "clawdbot": {
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/memory-core",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot core memory search plugin",
  "clawdbot": {
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/memory-lancedb",
-  "version": "0.0.1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot LanceDB-backed long-term memory plugin with auto-recall/capture",
  "dependencies": {
@@ -1,5 +1,15 @@
 # Changelog

+## 2026.1.20-2
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
+## 2026.1.20
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.17-1

 ### Changes
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/msteams",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Microsoft Teams channel plugin",
  "clawdbot": {
@@ -14,7 +14,9 @@
      "docsPath": "/channels/msteams",
      "docsLabel": "msteams",
      "blurb": "Bot Framework; enterprise support.",
-      "aliases": ["teams"],
+      "aliases": [
+        "teams"
+      ],
      "order": 60
    },
    "install": {
@@ -1,10 +1,12 @@
 {
  "name": "@clawdbot/nextcloud-talk",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Nextcloud Talk channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"],
+    "extensions": [
+      "./index.ts"
+    ],
    "channel": {
      "id": "nextcloud-talk",
      "label": "Nextcloud Talk",
@@ -12,7 +14,10 @@
      "docsPath": "/channels/nextcloud-talk",
      "docsLabel": "nextcloud-talk",
      "blurb": "Self-hosted chat via Nextcloud Talk webhook bots.",
-      "aliases": ["nc-talk", "nc"],
+      "aliases": [
+        "nc-talk",
+        "nc"
+      ],
      "order": 65,
      "quickstartAllowFrom": true
    },
@@ -1,5 +1,15 @@
 # Changelog

+## 2026.1.20-2
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
+## 2026.1.20
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.19-1

 Initial release.
@@ -1,10 +1,12 @@
 {
  "name": "@clawdbot/nostr",
-  "version": "2026.1.19-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Nostr channel plugin for NIP-04 encrypted DMs",
  "clawdbot": {
-    "extensions": ["./index.ts"],
+    "extensions": [
+      "./index.ts"
+    ],
    "channel": {
      "id": "nostr",
      "label": "Nostr",
@@ -1,9 +1,11 @@
 {
  "name": "@clawdbot/signal",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Signal channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  }
 }
@@ -1,9 +1,11 @@
 {
  "name": "@clawdbot/slack",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Slack channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  }
 }
@@ -1,9 +1,11 @@
 {
  "name": "@clawdbot/telegram",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Telegram channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  }
 }
@@ -1,5 +1,15 @@
 # Changelog

+## 2026.1.20-2
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
+## 2026.1.20
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.17-1

 ### Changes
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/voice-call",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot voice-call plugin",
  "dependencies": {
@@ -1,9 +1,11 @@
 {
  "name": "@clawdbot/whatsapp",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot WhatsApp channel plugin",
  "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
  }
 }
@@ -1,5 +1,15 @@
 # Changelog

+## 2026.1.20-2
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
+## 2026.1.20
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.17-1

 ### Changes
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/zalo",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Zalo channel plugin",
  "clawdbot": {
@@ -14,7 +14,9 @@
      "docsPath": "/channels/zalo",
      "docsLabel": "zalo",
      "blurb": "Vietnam-focused messaging platform with Bot API.",
-      "aliases": ["zl"],
+      "aliases": [
+        "zl"
+      ],
      "order": 80,
      "quickstartAllowFrom": true
    },
@@ -1,5 +1,15 @@
 # Changelog

+## 2026.1.20-2
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
+## 2026.1.20
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.17-1

 - Initial version with full channel plugin support
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/zalouser",
-  "version": "2026.1.17-1",
+  "version": "2026.1.20-2",
  "type": "module",
  "description": "Clawdbot Zalo Personal Account plugin via zca-cli",
  "dependencies": {
@@ -1,6 +1,6 @@
 {
  "name": "clawdbot",
-  "version": "2026.1.20",
+  "version": "2026.1.20-2",
  "description": "WhatsApp gateway CLI (Baileys web) with Pi RPC agent",
  "type": "module",
  "main": "dist/index.js",
@@ -13,6 +13,7 @@
    "clawdbot": "dist/entry.js"
  },
  "files": [
+    "dist/acp/**",
    "dist/agents/**",
    "dist/auto-reply/**",
    "dist/browser/**",
@@ -50,6 +51,7 @@
    "extensions/**",
    "assets/**",
    "skills/**",
+    "patches/**",
    "README.md",
    "README-header.png",
    "CHANGELOG.md",
@@ -64,6 +66,7 @@
    "dist/logging/**",
    "dist/memory/**",
    "dist/markdown/**",
+    "dist/node-host/**",
    "dist/pairing/**",
    "dist/whatsapp/**"
  ],
@@ -0,0 +1,93 @@
+import { describe, expect, it } from "vitest";
+
+import type { ClawdbotConfig } from "../config/config.js";
+import { resolveUserPath } from "../utils.js";
+import { createCacheTrace } from "./cache-trace.js";
+
+describe("createCacheTrace", () => {
+  it("returns null when diagnostics cache tracing is disabled", () => {
+    const trace = createCacheTrace({
+      cfg: {} as ClawdbotConfig,
+      env: {},
+    });
+
+    expect(trace).toBeNull();
+  });
+
+  it("honors diagnostics cache trace config and expands file paths", () => {
+    const lines: string[] = [];
+    const trace = createCacheTrace({
+      cfg: {
+        diagnostics: {
+          cacheTrace: {
+            enabled: true,
+            filePath: "~/.clawdbot/logs/cache-trace.jsonl",
+          },
+        },
+      },
+      env: {},
+      writer: {
+        filePath: "memory",
+        write: (line) => lines.push(line),
+      },
+    });
+
+    expect(trace).not.toBeNull();
+    expect(trace?.filePath).toBe(resolveUserPath("~/.clawdbot/logs/cache-trace.jsonl"));
+
+    trace?.recordStage("session:loaded", {
+      messages: [],
+      system: "sys",
+    });
+
+    expect(lines.length).toBe(1);
+  });
+
+  it("records empty prompt/system values when enabled", () => {
+    const lines: string[] = [];
+    const trace = createCacheTrace({
+      cfg: {
+        diagnostics: {
+          cacheTrace: {
+            enabled: true,
+            includePrompt: true,
+            includeSystem: true,
+          },
+        },
+      },
+      env: {},
+      writer: {
+        filePath: "memory",
+        write: (line) => lines.push(line),
+      },
+    });
+
+    trace?.recordStage("prompt:before", { prompt: "", system: "" });
+
+    const event = JSON.parse(lines[0]?.trim() ?? "{}") as Record<string, unknown>;
+    expect(event.prompt).toBe("");
+    expect(event.system).toBe("");
+  });
+
+  it("respects env overrides for enablement", () => {
+    const lines: string[] = [];
+    const trace = createCacheTrace({
+      cfg: {
+        diagnostics: {
+          cacheTrace: {
+            enabled: true,
+          },
+        },
+      },
+      env: {
+        CLAWDBOT_CACHE_TRACE: "0",
+      },
+      writer: {
+        filePath: "memory",
+        write: (line) => lines.push(line),
+      },
+    });
+
+    expect(trace).toBeNull();
+  });
+});
@@ -0,0 +1,271 @@
+import crypto from "node:crypto";
+import fs from "node:fs/promises";
+import path from "node:path";
+
+import type { AgentMessage, StreamFn } from "@mariozechner/pi-agent-core";
+import type { Api, Model } from "@mariozechner/pi-ai";
+
+import type { ClawdbotConfig } from "../config/config.js";
+import { resolveStateDir } from "../config/paths.js";
+import { parseBooleanValue } from "../utils/boolean.js";
+import { resolveUserPath } from "../utils.js";
+
+export type CacheTraceStage =
+  | "session:loaded"
+  | "session:sanitized"
+  | "session:limited"
+  | "prompt:before"
+  | "prompt:images"
+  | "stream:context"
+  | "session:after";
+
+export type CacheTraceEvent = {
+  ts: string;
+  seq: number;
+  stage: CacheTraceStage;
+  runId?: string;
+  sessionId?: string;
+  sessionKey?: string;
+  provider?: string;
+  modelId?: string;
+  modelApi?: string | null;
+  workspaceDir?: string;
+  prompt?: string;
+  system?: unknown;
+  options?: Record<string, unknown>;
+  model?: Record<string, unknown>;
+  messages?: AgentMessage[];
+  messageCount?: number;
+  messageRoles?: Array<string | undefined>;
+  messageFingerprints?: string[];
+  messagesDigest?: string;
+  systemDigest?: string;
+  note?: string;
+  error?: string;
+};
+
+export type CacheTrace = {
+  enabled: true;
+  filePath: string;
+  recordStage: (stage: CacheTraceStage, payload?: Partial<CacheTraceEvent>) => void;
+  wrapStreamFn: (streamFn: StreamFn) => StreamFn;
+};
+
+type CacheTraceInit = {
+  cfg?: ClawdbotConfig;
+  env?: NodeJS.ProcessEnv;
+  runId?: string;
+  sessionId?: string;
+  sessionKey?: string;
+  provider?: string;
+  modelId?: string;
+  modelApi?: string | null;
+  workspaceDir?: string;
+  writer?: CacheTraceWriter;
+};
+
+type CacheTraceConfig = {
+  enabled: boolean;
+  filePath: string;
+  includeMessages: boolean;
+  includePrompt: boolean;
+  includeSystem: boolean;
+};
+
+type CacheTraceWriter = {
+  filePath: string;
+  write: (line: string) => void;
+};
+
+const writers = new Map<string, CacheTraceWriter>();
+
+function resolveCacheTraceConfig(params: CacheTraceInit): CacheTraceConfig {
+  const env = params.env ?? process.env;
+  const config = params.cfg?.diagnostics?.cacheTrace;
+  const envEnabled = parseBooleanValue(env.CLAWDBOT_CACHE_TRACE);
+  const enabled = envEnabled ?? config?.enabled ?? false;
+  const fileOverride = config?.filePath?.trim() || env.CLAWDBOT_CACHE_TRACE_FILE?.trim();
+  const filePath = fileOverride
+    ? resolveUserPath(fileOverride)
+    : path.join(resolveStateDir(env), "logs", "cache-trace.jsonl");
+
+  const includeMessages =
+    parseBooleanValue(env.CLAWDBOT_CACHE_TRACE_MESSAGES) ?? config?.includeMessages;
+  const includePrompt = parseBooleanValue(env.CLAWDBOT_CACHE_TRACE_PROMPT) ?? config?.includePrompt;
+  const includeSystem = parseBooleanValue(env.CLAWDBOT_CACHE_TRACE_SYSTEM) ?? config?.includeSystem;
+
+  return {
+    enabled,
+    filePath,
+    includeMessages: includeMessages ?? true,
+    includePrompt: includePrompt ?? true,
+    includeSystem: includeSystem ?? true,
+  };
+}
+
+function getWriter(filePath: string): CacheTraceWriter {
+  const existing = writers.get(filePath);
+  if (existing) return existing;
+
+  const dir = path.dirname(filePath);
+  const ready = fs.mkdir(dir, { recursive: true }).catch(() => undefined);
+  let queue = Promise.resolve();
+
+  const writer: CacheTraceWriter = {
+    filePath,
+    write: (line: string) => {
+      queue = queue
+        .then(() => ready)
+        .then(() => fs.appendFile(filePath, line, "utf8"))
+        .catch(() => undefined);
+    },
+  };
+
+  writers.set(filePath, writer);
+  return writer;
+}
+
+function stableStringify(value: unknown): string {
+  if (value === null || value === undefined) return String(value);
+  if (typeof value === "number" && !Number.isFinite(value)) return JSON.stringify(String(value));
+  if (typeof value === "bigint") return JSON.stringify(value.toString());
+  if (typeof value !== "object") return JSON.stringify(value) ?? "null";
+  if (value instanceof Error) {
+    return stableStringify({
+      name: value.name,
+      message: value.message,
+      stack: value.stack,
+    });
+  }
+  if (value instanceof Uint8Array) {
+    return stableStringify({
+      type: "Uint8Array",
+      data: Buffer.from(value).toString("base64"),
+    });
+  }
+  if (Array.isArray(value)) {
+    return `[${value.map((entry) => stableStringify(entry)).join(",")}]`;
+  }
+  const record = value as Record<string, unknown>;
+  const keys = Object.keys(record).sort();
+  const entries = keys.map((key) => `${JSON.stringify(key)}:${stableStringify(record[key])}`);
+  return `{${entries.join(",")}}`;
+}
+
+function digest(value: unknown): string {
+  const serialized = stableStringify(value);
+  return crypto.createHash("sha256").update(serialized).digest("hex");
+}
+
+function summarizeMessages(messages: AgentMessage[]): {
+  messageCount: number;
+  messageRoles: Array<string | undefined>;
+  messageFingerprints: string[];
+  messagesDigest: string;
+} {
+  const messageFingerprints = messages.map((msg) => digest(msg));
+  return {
+    messageCount: messages.length,
+    messageRoles: messages.map((msg) => (msg as { role?: string }).role),
+    messageFingerprints,
+    messagesDigest: digest(messageFingerprints.join("|")),
+  };
+}
+
+function safeJsonStringify(value: unknown): string | null {
+  try {
+    return JSON.stringify(value, (_key, val) => {
+      if (typeof val === "bigint") return val.toString();
+      if (typeof val === "function") return "[Function]";
+      if (val instanceof Error) {
+        return { name: val.name, message: val.message, stack: val.stack };
+      }
+      if (val instanceof Uint8Array) {
+        return { type: "Uint8Array", data: Buffer.from(val).toString("base64") };
+      }
+      return val;
+    });
+  } catch {
+    return null;
+  }
+}
+
+export function createCacheTrace(params: CacheTraceInit): CacheTrace | null {
+  const cfg = resolveCacheTraceConfig(params);
+  if (!cfg.enabled) return null;
+
+  const writer = params.writer ?? getWriter(cfg.filePath);
+  let seq = 0;
+
+  const base: Omit<CacheTraceEvent, "ts" | "seq" | "stage"> = {
+    runId: params.runId,
+    sessionId: params.sessionId,
+    sessionKey: params.sessionKey,
+    provider: params.provider,
+    modelId: params.modelId,
+    modelApi: params.modelApi,
+    workspaceDir: params.workspaceDir,
+  };
+
+  const recordStage: CacheTrace["recordStage"] = (stage, payload = {}) => {
+    const event: CacheTraceEvent = {
+      ...base,
+      ts: new Date().toISOString(),
+      seq: (seq += 1),
+      stage,
+    };
+
+    if (payload.prompt !== undefined && cfg.includePrompt) {
+      event.prompt = payload.prompt;
+    }
+    if (payload.system !== undefined && cfg.includeSystem) {
+      event.system = payload.system;
+      event.systemDigest = digest(payload.system);
+    }
+    if (payload.options) event.options = payload.options;
+    if (payload.model) event.model = payload.model;
+
+    const messages = payload.messages;
+    if (Array.isArray(messages)) {
+      const summary = summarizeMessages(messages);
+      event.messageCount = summary.messageCount;
+      event.messageRoles = summary.messageRoles;
+      event.messageFingerprints = summary.messageFingerprints;
+      event.messagesDigest = summary.messagesDigest;
+      if (cfg.includeMessages) {
+        event.messages = messages;
+      }
+    }
+
+    if (payload.note) event.note = payload.note;
+    if (payload.error) event.error = payload.error;
+
+    const line = safeJsonStringify(event);
+    if (!line) return;
+    writer.write(`${line}\n`);
+  };
+
+  const wrapStreamFn: CacheTrace["wrapStreamFn"] = (streamFn) => {
+    const wrapped: StreamFn = (model, context, options) => {
+      recordStage("stream:context", {
+        model: {
+          id: (model as Model<Api>)?.id,
+          provider: (model as Model<Api>)?.provider,
+          api: (model as Model<Api>)?.api,
+        },
+        system: (context as { system?: unknown }).system,
+        messages: (context as { messages?: AgentMessage[] }).messages ?? [],
+        options: (options ?? {}) as Record<string, unknown>,
+      });
+      return streamFn(model, context, options);
+    };
+    return wrapped;
+  };
+
+  return {
+    enabled: true,
+    filePath: cfg.filePath,
+    recordStage,
+    wrapStreamFn,
+  };
+}
@@ -4,6 +4,9 @@ const callGatewayMock = vi.fn();
 vi.mock("../gateway/call.js", () => ({
  callGateway: (opts: unknown) => callGatewayMock(opts),
 }));
+vi.mock("../plugins/tools.js", () => ({
+  resolvePluginTools: () => [],
+}));

 vi.mock("../config/config.js", async (importOriginal) => {
  const actual = await importOriginal<typeof import("../config/config.js")>();
@@ -136,6 +136,23 @@ describe("memory search config", () => {
    });
  });

+  it("defaults session delta thresholds", () => {
+    const cfg = {
+      agents: {
+        defaults: {
+          memorySearch: {
+            provider: "openai",
+          },
+        },
+      },
+    };
+    const resolved = resolveMemorySearchConfig(cfg, "main");
+    expect(resolved?.sync.sessions).toEqual({
+      deltaBytes: 100000,
+      deltaMessages: 50,
+    });
+  });
+
  it("merges remote defaults with agent overrides", () => {
    const cfg = {
      agents: {
@@ -49,6 +49,10 @@ export type ResolvedMemorySearchConfig = {
    watch: boolean;
    watchDebounceMs: number;
    intervalMinutes: number;
+    sessions: {
+      deltaBytes: number;
+      deltaMessages: number;
+    };
  };
  query: {
    maxResults: number;
@@ -71,6 +75,8 @@ const DEFAULT_GEMINI_MODEL = "gemini-embedding-001";
 const DEFAULT_CHUNK_TOKENS = 400;
 const DEFAULT_CHUNK_OVERLAP = 80;
 const DEFAULT_WATCH_DEBOUNCE_MS = 1500;
+const DEFAULT_SESSION_DELTA_BYTES = 100_000;
+const DEFAULT_SESSION_DELTA_MESSAGES = 50;
 const DEFAULT_MAX_RESULTS = 6;
 const DEFAULT_MIN_SCORE = 0.35;
 const DEFAULT_HYBRID_ENABLED = true;
@@ -172,6 +178,16 @@ function mergeConfig(
      defaults?.sync?.watchDebounceMs ??
      DEFAULT_WATCH_DEBOUNCE_MS,
    intervalMinutes: overrides?.sync?.intervalMinutes ?? defaults?.sync?.intervalMinutes ?? 0,
+    sessions: {
+      deltaBytes:
+        overrides?.sync?.sessions?.deltaBytes ??
+        defaults?.sync?.sessions?.deltaBytes ??
+        DEFAULT_SESSION_DELTA_BYTES,
+      deltaMessages:
+        overrides?.sync?.sessions?.deltaMessages ??
+        defaults?.sync?.sessions?.deltaMessages ??
+        DEFAULT_SESSION_DELTA_MESSAGES,
+    },
  };
  const query = {
    maxResults: overrides?.query?.maxResults ?? defaults?.query?.maxResults ?? DEFAULT_MAX_RESULTS,
@@ -208,6 +224,8 @@ function mergeConfig(
  const normalizedVectorWeight = sum > 0 ? vectorWeight / sum : DEFAULT_HYBRID_VECTOR_WEIGHT;
  const normalizedTextWeight = sum > 0 ? textWeight / sum : DEFAULT_HYBRID_TEXT_WEIGHT;
  const candidateMultiplier = clampInt(hybrid.candidateMultiplier, 1, 20);
+  const deltaBytes = clampInt(sync.sessions.deltaBytes, 0, Number.MAX_SAFE_INTEGER);
+  const deltaMessages = clampInt(sync.sessions.deltaMessages, 0, Number.MAX_SAFE_INTEGER);
  return {
    enabled,
    sources,
@@ -221,7 +239,13 @@ function mergeConfig(
    local,
    store,
    chunking: { tokens: Math.max(1, chunking.tokens), overlap },
-    sync,
+    sync: {
+      ...sync,
+      sessions: {
+        deltaBytes,
+        deltaMessages,
+      },
+    },
    query: {
      ...query,
      minScore,
@@ -51,6 +51,18 @@ import { describeUnknownError } from "./utils.js";

 type ApiKeyInfo = ResolvedProviderAuth;

+// Avoid Anthropic's refusal test token poisoning session transcripts.
+const ANTHROPIC_MAGIC_STRING_TRIGGER_REFUSAL = "ANTHROPIC_MAGIC_STRING_TRIGGER_REFUSAL";
+const ANTHROPIC_MAGIC_STRING_REPLACEMENT = "ANTHROPIC MAGIC STRING TRIGGER REFUSAL (redacted)";
+
+function scrubAnthropicRefusalMagic(prompt: string): string {
+  if (!prompt.includes(ANTHROPIC_MAGIC_STRING_TRIGGER_REFUSAL)) return prompt;
+  return prompt.replaceAll(
+    ANTHROPIC_MAGIC_STRING_TRIGGER_REFUSAL,
+    ANTHROPIC_MAGIC_STRING_REPLACEMENT,
+  );
+}
+
 export async function runEmbeddedPiAgent(
  params: RunEmbeddedPiAgentParams,
 ): Promise<EmbeddedPiRunResult> {
@@ -211,6 +223,9 @@ export async function runEmbeddedPiAgent(
          attemptedThinking.add(thinkLevel);
          await fs.mkdir(resolvedWorkspace, { recursive: true });

+          const prompt =
+            provider === "anthropic" ? scrubAnthropicRefusalMagic(params.prompt) : params.prompt;
+
          const attempt = await runEmbeddedAttempt({
            sessionId: params.sessionId,
            sessionKey: params.sessionKey,
@@ -228,7 +243,7 @@ export async function runEmbeddedPiAgent(
            agentDir,
            config: params.config,
            skillsSnapshot: params.skillsSnapshot,
-            prompt: params.prompt,
+            prompt,
            images: params.images,
            provider,
            modelId,
@@ -16,6 +16,7 @@ import { normalizeMessageChannel } from "../../../utils/message-channel.js";
 import { isReasoningTagProvider } from "../../../utils/provider-utils.js";
 import { isSubagentSessionKey } from "../../../routing/session-key.js";
 import { resolveUserPath } from "../../../utils.js";
+import { createCacheTrace } from "../../cache-trace.js";
 import { resolveClawdbotAgentDir } from "../../agent-paths.js";
 import { resolveSessionAgentIds } from "../../agent-scope.js";
 import { makeBootstrapWarn, resolveBootstrapContextForRun } from "../../bootstrap-files.js";
@@ -367,6 +368,17 @@ export async function runEmbeddedAttempt(
        throw new Error("Embedded agent session missing");
      }
      const activeSession = session;
+      const cacheTrace = createCacheTrace({
+        cfg: params.config,
+        env: process.env,
+        runId: params.runId,
+        sessionId: activeSession.sessionId,
+        sessionKey: params.sessionKey,
+        provider: params.provider,
+        modelId: params.modelId,
+        modelApi: params.model.api,
+        workspaceDir: params.workspaceDir,
+      });

      // Force a stable streamFn reference so vitest can reliably mock @mariozechner/pi-ai.
      activeSession.agent.streamFn = streamSimple;
@@ -379,6 +391,15 @@ export async function runEmbeddedAttempt(
        params.streamParams,
      );

+      if (cacheTrace) {
+        cacheTrace.recordStage("session:loaded", {
+          messages: activeSession.messages,
+          system: systemPrompt,
+          note: "after session create",
+        });
+        activeSession.agent.streamFn = cacheTrace.wrapStreamFn(activeSession.agent.streamFn);
+      }
+
      try {
        const prior = await sanitizeSessionHistory({
          messages: activeSession.messages,
@@ -388,12 +409,14 @@ export async function runEmbeddedAttempt(
          sessionManager,
          sessionId: params.sessionId,
        });
+        cacheTrace?.recordStage("session:sanitized", { messages: prior });
        const validatedGemini = validateGeminiTurns(prior);
        const validated = validateAnthropicTurns(validatedGemini);
        const limited = limitHistoryTurns(
          validated,
          getDmHistoryLimitFromSessionKey(params.sessionKey, params.config),
        );
+        cacheTrace?.recordStage("session:limited", { messages: limited });
        if (limited.length > 0) {
          activeSession.agent.replaceMessages(limited);
        }
@@ -564,6 +587,10 @@ export async function runEmbeddedAttempt(
        }

        log.debug(`embedded run prompt start: runId=${params.runId} sessionId=${params.sessionId}`);
+        cacheTrace?.recordStage("prompt:before", {
+          prompt: effectivePrompt,
+          messages: activeSession.messages,
+        });

        // Repair orphaned trailing user messages so new prompts don't violate role ordering.
        const leafEntry = sessionManager.getLeafEntry();
@@ -633,6 +660,12 @@ export async function runEmbeddedAttempt(
            }
          }

+          cacheTrace?.recordStage("prompt:images", {
+            prompt: effectivePrompt,
+            messages: activeSession.messages,
+            note: `images: prompt=${imageResult.images.length} history=${imageResult.historyImagesByIndex.size}`,
+          });
+
          // Only pass images option if there are actually images to pass
          // This avoids potential issues with models that don't expect the images parameter
          if (imageResult.images.length > 0) {
@@ -660,6 +693,10 @@ export async function runEmbeddedAttempt(

        messagesSnapshot = activeSession.messages.slice();
        sessionIdUsed = activeSession.sessionId;
+        cacheTrace?.recordStage("session:after", {
+          messages: messagesSnapshot,
+          note: promptError ? "prompt error" : undefined,
+        });

        // Run agent_end hooks to allow plugins to analyze the conversation
        // This is fire-and-forget, so we don't await
@@ -112,334 +112,349 @@ export function createNodesTool(options?: {
        timeoutMs: typeof params.timeoutMs === "number" ? params.timeoutMs : undefined,
      };

-      switch (action) {
-        case "status":
-          return jsonResult(await callGatewayTool("node.list", gatewayOpts, {}));
-        case "describe": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          return jsonResult(await callGatewayTool("node.describe", gatewayOpts, { nodeId }));
-        }
-        case "pending":
-          return jsonResult(await callGatewayTool("node.pair.list", gatewayOpts, {}));
-        case "approve": {
-          const requestId = readStringParam(params, "requestId", {
-            required: true,
-          });
-          return jsonResult(
-            await callGatewayTool("node.pair.approve", gatewayOpts, {
-              requestId,
-            }),
-          );
-        }
-        case "reject": {
-          const requestId = readStringParam(params, "requestId", {
-            required: true,
-          });
-          return jsonResult(
-            await callGatewayTool("node.pair.reject", gatewayOpts, {
-              requestId,
-            }),
-          );
-        }
-        case "notify": {
-          const node = readStringParam(params, "node", { required: true });
-          const title = typeof params.title === "string" ? params.title : "";
-          const body = typeof params.body === "string" ? params.body : "";
-          if (!title.trim() && !body.trim()) {
-            throw new Error("title or body required");
+      try {
+        switch (action) {
+          case "status":
+            return jsonResult(await callGatewayTool("node.list", gatewayOpts, {}));
+          case "describe": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodeId = await resolveNodeId(gatewayOpts, node);
+            return jsonResult(await callGatewayTool("node.describe", gatewayOpts, { nodeId }));
          }
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          await callGatewayTool("node.invoke", gatewayOpts, {
-            nodeId,
-            command: "system.notify",
-            params: {
-              title: title.trim() || undefined,
-              body: body.trim() || undefined,
-              sound: typeof params.sound === "string" ? params.sound : undefined,
-              priority: typeof params.priority === "string" ? params.priority : undefined,
-              delivery: typeof params.delivery === "string" ? params.delivery : undefined,
-            },
-            idempotencyKey: crypto.randomUUID(),
-          });
-          return jsonResult({ ok: true });
-        }
-        case "camera_snap": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          const facingRaw =
-            typeof params.facing === "string" ? params.facing.toLowerCase() : "both";
-          const facings: CameraFacing[] =
-            facingRaw === "both"
-              ? ["front", "back"]
-              : facingRaw === "front" || facingRaw === "back"
-                ? [facingRaw]
-                : (() => {
-                    throw new Error("invalid facing (front|back|both)");
-                  })();
-          const maxWidth =
-            typeof params.maxWidth === "number" && Number.isFinite(params.maxWidth)
-              ? params.maxWidth
-              : undefined;
-          const quality =
-            typeof params.quality === "number" && Number.isFinite(params.quality)
-              ? params.quality
-              : undefined;
-          const delayMs =
-            typeof params.delayMs === "number" && Number.isFinite(params.delayMs)
-              ? params.delayMs
-              : undefined;
-          const deviceId =
-            typeof params.deviceId === "string" && params.deviceId.trim()
-              ? params.deviceId.trim()
-              : undefined;
+          case "pending":
+            return jsonResult(await callGatewayTool("node.pair.list", gatewayOpts, {}));
+          case "approve": {
+            const requestId = readStringParam(params, "requestId", {
+              required: true,
+            });
+            return jsonResult(
+              await callGatewayTool("node.pair.approve", gatewayOpts, {
+                requestId,
+              }),
+            );
+          }
+          case "reject": {
+            const requestId = readStringParam(params, "requestId", {
+              required: true,
+            });
+            return jsonResult(
+              await callGatewayTool("node.pair.reject", gatewayOpts, {
+                requestId,
+              }),
+            );
+          }
+          case "notify": {
+            const node = readStringParam(params, "node", { required: true });
+            const title = typeof params.title === "string" ? params.title : "";
+            const body = typeof params.body === "string" ? params.body : "";
+            if (!title.trim() && !body.trim()) {
+              throw new Error("title or body required");
+            }
+            const nodeId = await resolveNodeId(gatewayOpts, node);
+            await callGatewayTool("node.invoke", gatewayOpts, {
+              nodeId,
+              command: "system.notify",
+              params: {
+                title: title.trim() || undefined,
+                body: body.trim() || undefined,
+                sound: typeof params.sound === "string" ? params.sound : undefined,
+                priority: typeof params.priority === "string" ? params.priority : undefined,
+                delivery: typeof params.delivery === "string" ? params.delivery : undefined,
+              },
+              idempotencyKey: crypto.randomUUID(),
+            });
+            return jsonResult({ ok: true });
+          }
+          case "camera_snap": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodeId = await resolveNodeId(gatewayOpts, node);
+            const facingRaw =
+              typeof params.facing === "string" ? params.facing.toLowerCase() : "both";
+            const facings: CameraFacing[] =
+              facingRaw === "both"
+                ? ["front", "back"]
+                : facingRaw === "front" || facingRaw === "back"
+                  ? [facingRaw]
+                  : (() => {
+                      throw new Error("invalid facing (front|back|both)");
+                    })();
+            const maxWidth =
+              typeof params.maxWidth === "number" && Number.isFinite(params.maxWidth)
+                ? params.maxWidth
+                : undefined;
+            const quality =
+              typeof params.quality === "number" && Number.isFinite(params.quality)
+                ? params.quality
+                : undefined;
+            const delayMs =
+              typeof params.delayMs === "number" && Number.isFinite(params.delayMs)
+                ? params.delayMs
+                : undefined;
+            const deviceId =
+              typeof params.deviceId === "string" && params.deviceId.trim()
+                ? params.deviceId.trim()
+                : undefined;

-          const content: AgentToolResult<unknown>["content"] = [];
-          const details: Array<Record<string, unknown>> = [];
+            const content: AgentToolResult<unknown>["content"] = [];
+            const details: Array<Record<string, unknown>> = [];

-          for (const facing of facings) {
+            for (const facing of facings) {
+              const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
+                nodeId,
+                command: "camera.snap",
+                params: {
+                  facing,
+                  maxWidth,
+                  quality,
+                  format: "jpg",
+                  delayMs,
+                  deviceId,
+                },
+                idempotencyKey: crypto.randomUUID(),
+              })) as { payload?: unknown };
+              const payload = parseCameraSnapPayload(raw?.payload);
+              const normalizedFormat = payload.format.toLowerCase();
+              if (
+                normalizedFormat !== "jpg" &&
+                normalizedFormat !== "jpeg" &&
+                normalizedFormat !== "png"
+              ) {
+                throw new Error(`unsupported camera.snap format: ${payload.format}`);
+              }
+
+              const isJpeg = normalizedFormat === "jpg" || normalizedFormat === "jpeg";
+              const filePath = cameraTempPath({
+                kind: "snap",
+                facing,
+                ext: isJpeg ? "jpg" : "png",
+              });
+              await writeBase64ToFile(filePath, payload.base64);
+              content.push({ type: "text", text: `MEDIA:${filePath}` });
+              content.push({
+                type: "image",
+                data: payload.base64,
+                mimeType:
+                  imageMimeFromFormat(payload.format) ?? (isJpeg ? "image/jpeg" : "image/png"),
+              });
+              details.push({
+                facing,
+                path: filePath,
+                width: payload.width,
+                height: payload.height,
+              });
+            }
+
+            const result: AgentToolResult<unknown> = { content, details };
+            return await sanitizeToolResultImages(result, "nodes:camera_snap");
+          }
+          case "camera_list": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodeId = await resolveNodeId(gatewayOpts, node);
            const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
              nodeId,
-              command: "camera.snap",
+              command: "camera.list",
+              params: {},
+              idempotencyKey: crypto.randomUUID(),
+            })) as { payload?: unknown };
+            const payload =
+              raw && typeof raw.payload === "object" && raw.payload !== null ? raw.payload : {};
+            return jsonResult(payload);
+          }
+          case "camera_clip": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodeId = await resolveNodeId(gatewayOpts, node);
+            const facing =
+              typeof params.facing === "string" ? params.facing.toLowerCase() : "front";
+            if (facing !== "front" && facing !== "back") {
+              throw new Error("invalid facing (front|back)");
+            }
+            const durationMs =
+              typeof params.durationMs === "number" && Number.isFinite(params.durationMs)
+                ? params.durationMs
+                : typeof params.duration === "string"
+                  ? parseDurationMs(params.duration)
+                  : 3000;
+            const includeAudio =
+              typeof params.includeAudio === "boolean" ? params.includeAudio : true;
+            const deviceId =
+              typeof params.deviceId === "string" && params.deviceId.trim()
+                ? params.deviceId.trim()
+                : undefined;
+            const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
+              nodeId,
+              command: "camera.clip",
              params: {
                facing,
-                maxWidth,
-                quality,
-                format: "jpg",
-                delayMs,
+                durationMs,
+                includeAudio,
+                format: "mp4",
                deviceId,
              },
              idempotencyKey: crypto.randomUUID(),
            })) as { payload?: unknown };
-            const payload = parseCameraSnapPayload(raw?.payload);
-            const normalizedFormat = payload.format.toLowerCase();
-            if (
-              normalizedFormat !== "jpg" &&
-              normalizedFormat !== "jpeg" &&
-              normalizedFormat !== "png"
-            ) {
-              throw new Error(`unsupported camera.snap format: ${payload.format}`);
-            }
-
-            const isJpeg = normalizedFormat === "jpg" || normalizedFormat === "jpeg";
+            const payload = parseCameraClipPayload(raw?.payload);
            const filePath = cameraTempPath({
-              kind: "snap",
+              kind: "clip",
              facing,
-              ext: isJpeg ? "jpg" : "png",
+              ext: payload.format,
            });
            await writeBase64ToFile(filePath, payload.base64);
-            content.push({ type: "text", text: `MEDIA:${filePath}` });
-            content.push({
-              type: "image",
-              data: payload.base64,
-              mimeType:
-                imageMimeFromFormat(payload.format) ?? (isJpeg ? "image/jpeg" : "image/png"),
-            });
-            details.push({
-              facing,
-              path: filePath,
-              width: payload.width,
-              height: payload.height,
-            });
+            return {
+              content: [{ type: "text", text: `FILE:${filePath}` }],
+              details: {
+                facing,
+                path: filePath,
+                durationMs: payload.durationMs,
+                hasAudio: payload.hasAudio,
+              },
+            };
          }
-
-          const result: AgentToolResult<unknown> = { content, details };
-          return await sanitizeToolResultImages(result, "nodes:camera_snap");
+          case "screen_record": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodeId = await resolveNodeId(gatewayOpts, node);
+            const durationMs =
+              typeof params.durationMs === "number" && Number.isFinite(params.durationMs)
+                ? params.durationMs
+                : typeof params.duration === "string"
+                  ? parseDurationMs(params.duration)
+                  : 10_000;
+            const fps =
+              typeof params.fps === "number" && Number.isFinite(params.fps) ? params.fps : 10;
+            const screenIndex =
+              typeof params.screenIndex === "number" && Number.isFinite(params.screenIndex)
+                ? params.screenIndex
+                : 0;
+            const includeAudio =
+              typeof params.includeAudio === "boolean" ? params.includeAudio : true;
+            const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
+              nodeId,
+              command: "screen.record",
+              params: {
+                durationMs,
+                screenIndex,
+                fps,
+                format: "mp4",
+                includeAudio,
+              },
+              idempotencyKey: crypto.randomUUID(),
+            })) as { payload?: unknown };
+            const payload = parseScreenRecordPayload(raw?.payload);
+            const filePath =
+              typeof params.outPath === "string" && params.outPath.trim()
+                ? params.outPath.trim()
+                : screenRecordTempPath({ ext: payload.format || "mp4" });
+            const written = await writeScreenRecordToFile(filePath, payload.base64);
+            return {
+              content: [{ type: "text", text: `FILE:${written.path}` }],
+              details: {
+                path: written.path,
+                durationMs: payload.durationMs,
+                fps: payload.fps,
+                screenIndex: payload.screenIndex,
+                hasAudio: payload.hasAudio,
+              },
+            };
+          }
+          case "location_get": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodeId = await resolveNodeId(gatewayOpts, node);
+            const maxAgeMs =
+              typeof params.maxAgeMs === "number" && Number.isFinite(params.maxAgeMs)
+                ? params.maxAgeMs
+                : undefined;
+            const desiredAccuracy =
+              params.desiredAccuracy === "coarse" ||
+              params.desiredAccuracy === "balanced" ||
+              params.desiredAccuracy === "precise"
+                ? params.desiredAccuracy
+                : undefined;
+            const locationTimeoutMs =
+              typeof params.locationTimeoutMs === "number" &&
+              Number.isFinite(params.locationTimeoutMs)
+                ? params.locationTimeoutMs
+                : undefined;
+            const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
+              nodeId,
+              command: "location.get",
+              params: {
+                maxAgeMs,
+                desiredAccuracy,
+                timeoutMs: locationTimeoutMs,
+              },
+              idempotencyKey: crypto.randomUUID(),
+            })) as { payload?: unknown };
+            return jsonResult(raw?.payload ?? {});
+          }
+          case "run": {
+            const node = readStringParam(params, "node", { required: true });
+            const nodes = await listNodes(gatewayOpts);
+            if (nodes.length === 0) {
+              throw new Error(
+                "system.run requires a paired companion app or node host (no nodes available).",
+              );
+            }
+            const nodeId = resolveNodeIdFromList(nodes, node);
+            const nodeInfo = nodes.find((entry) => entry.nodeId === nodeId);
+            const supportsSystemRun = Array.isArray(nodeInfo?.commands)
+              ? nodeInfo?.commands?.includes("system.run")
+              : false;
+            if (!supportsSystemRun) {
+              throw new Error(
+                "system.run requires a companion app or node host; the selected node does not support system.run.",
+              );
+            }
+            const commandRaw = params.command;
+            if (!commandRaw) {
+              throw new Error("command required (argv array, e.g. ['echo', 'Hello'])");
+            }
+            if (!Array.isArray(commandRaw)) {
+              throw new Error("command must be an array of strings (argv), e.g. ['echo', 'Hello']");
+            }
+            const command = commandRaw.map((c) => String(c));
+            if (command.length === 0) {
+              throw new Error("command must not be empty");
+            }
+            const cwd =
+              typeof params.cwd === "string" && params.cwd.trim() ? params.cwd.trim() : undefined;
+            const env = parseEnvPairs(params.env);
+            const commandTimeoutMs = parseTimeoutMs(params.commandTimeoutMs);
+            const invokeTimeoutMs = parseTimeoutMs(params.invokeTimeoutMs);
+            const needsScreenRecording =
+              typeof params.needsScreenRecording === "boolean"
+                ? params.needsScreenRecording
+                : undefined;
+            const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
+              nodeId,
+              command: "system.run",
+              params: {
+                command,
+                cwd,
+                env,
+                timeoutMs: commandTimeoutMs,
+                needsScreenRecording,
+                agentId,
+                sessionKey,
+              },
+              timeoutMs: invokeTimeoutMs,
+              idempotencyKey: crypto.randomUUID(),
+            })) as { payload?: unknown };
+            return jsonResult(raw?.payload ?? {});
+          }
+          default:
+            throw new Error(`Unknown action: ${action}`);
        }
-        case "camera_list": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
-            nodeId,
-            command: "camera.list",
-            params: {},
-            idempotencyKey: crypto.randomUUID(),
-          })) as { payload?: unknown };
-          const payload =
-            raw && typeof raw.payload === "object" && raw.payload !== null ? raw.payload : {};
-          return jsonResult(payload);
-        }
-        case "camera_clip": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          const facing = typeof params.facing === "string" ? params.facing.toLowerCase() : "front";
-          if (facing !== "front" && facing !== "back") {
-            throw new Error("invalid facing (front|back)");
-          }
-          const durationMs =
-            typeof params.durationMs === "number" && Number.isFinite(params.durationMs)
-              ? params.durationMs
-              : typeof params.duration === "string"
-                ? parseDurationMs(params.duration)
-                : 3000;
-          const includeAudio =
-            typeof params.includeAudio === "boolean" ? params.includeAudio : true;
-          const deviceId =
-            typeof params.deviceId === "string" && params.deviceId.trim()
-              ? params.deviceId.trim()
-              : undefined;
-          const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
-            nodeId,
-            command: "camera.clip",
-            params: {
-              facing,
-              durationMs,
-              includeAudio,
-              format: "mp4",
-              deviceId,
-            },
-            idempotencyKey: crypto.randomUUID(),
-          })) as { payload?: unknown };
-          const payload = parseCameraClipPayload(raw?.payload);
-          const filePath = cameraTempPath({
-            kind: "clip",
-            facing,
-            ext: payload.format,
-          });
-          await writeBase64ToFile(filePath, payload.base64);
-          return {
-            content: [{ type: "text", text: `FILE:${filePath}` }],
-            details: {
-              facing,
-              path: filePath,
-              durationMs: payload.durationMs,
-              hasAudio: payload.hasAudio,
-            },
-          };
-        }
-        case "screen_record": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          const durationMs =
-            typeof params.durationMs === "number" && Number.isFinite(params.durationMs)
-              ? params.durationMs
-              : typeof params.duration === "string"
-                ? parseDurationMs(params.duration)
-                : 10_000;
-          const fps =
-            typeof params.fps === "number" && Number.isFinite(params.fps) ? params.fps : 10;
-          const screenIndex =
-            typeof params.screenIndex === "number" && Number.isFinite(params.screenIndex)
-              ? params.screenIndex
-              : 0;
-          const includeAudio =
-            typeof params.includeAudio === "boolean" ? params.includeAudio : true;
-          const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
-            nodeId,
-            command: "screen.record",
-            params: {
-              durationMs,
-              screenIndex,
-              fps,
-              format: "mp4",
-              includeAudio,
-            },
-            idempotencyKey: crypto.randomUUID(),
-          })) as { payload?: unknown };
-          const payload = parseScreenRecordPayload(raw?.payload);
-          const filePath =
-            typeof params.outPath === "string" && params.outPath.trim()
-              ? params.outPath.trim()
-              : screenRecordTempPath({ ext: payload.format || "mp4" });
-          const written = await writeScreenRecordToFile(filePath, payload.base64);
-          return {
-            content: [{ type: "text", text: `FILE:${written.path}` }],
-            details: {
-              path: written.path,
-              durationMs: payload.durationMs,
-              fps: payload.fps,
-              screenIndex: payload.screenIndex,
-              hasAudio: payload.hasAudio,
-            },
-          };
-        }
-        case "location_get": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodeId = await resolveNodeId(gatewayOpts, node);
-          const maxAgeMs =
-            typeof params.maxAgeMs === "number" && Number.isFinite(params.maxAgeMs)
-              ? params.maxAgeMs
-              : undefined;
-          const desiredAccuracy =
-            params.desiredAccuracy === "coarse" ||
-            params.desiredAccuracy === "balanced" ||
-            params.desiredAccuracy === "precise"
-              ? params.desiredAccuracy
-              : undefined;
-          const locationTimeoutMs =
-            typeof params.locationTimeoutMs === "number" &&
-            Number.isFinite(params.locationTimeoutMs)
-              ? params.locationTimeoutMs
-              : undefined;
-          const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
-            nodeId,
-            command: "location.get",
-            params: {
-              maxAgeMs,
-              desiredAccuracy,
-              timeoutMs: locationTimeoutMs,
-            },
-            idempotencyKey: crypto.randomUUID(),
-          })) as { payload?: unknown };
-          return jsonResult(raw?.payload ?? {});
-        }
-        case "run": {
-          const node = readStringParam(params, "node", { required: true });
-          const nodes = await listNodes(gatewayOpts);
-          if (nodes.length === 0) {
-            throw new Error(
-              "system.run requires a paired companion app or node host (no nodes available).",
-            );
-          }
-          const nodeId = resolveNodeIdFromList(nodes, node);
-          const nodeInfo = nodes.find((entry) => entry.nodeId === nodeId);
-          const supportsSystemRun = Array.isArray(nodeInfo?.commands)
-            ? nodeInfo?.commands?.includes("system.run")
-            : false;
-          if (!supportsSystemRun) {
-            throw new Error(
-              "system.run requires a companion app or node host; the selected node does not support system.run.",
-            );
-          }
-          const commandRaw = params.command;
-          if (!commandRaw) {
-            throw new Error("command required (argv array, e.g. ['echo', 'Hello'])");
-          }
-          if (!Array.isArray(commandRaw)) {
-            throw new Error("command must be an array of strings (argv), e.g. ['echo', 'Hello']");
-          }
-          const command = commandRaw.map((c) => String(c));
-          if (command.length === 0) {
-            throw new Error("command must not be empty");
-          }
-          const cwd =
-            typeof params.cwd === "string" && params.cwd.trim() ? params.cwd.trim() : undefined;
-          const env = parseEnvPairs(params.env);
-          const commandTimeoutMs = parseTimeoutMs(params.commandTimeoutMs);
-          const invokeTimeoutMs = parseTimeoutMs(params.invokeTimeoutMs);
-          const needsScreenRecording =
-            typeof params.needsScreenRecording === "boolean"
-              ? params.needsScreenRecording
-              : undefined;
-          const raw = (await callGatewayTool("node.invoke", gatewayOpts, {
-            nodeId,
-            command: "system.run",
-            params: {
-              command,
-              cwd,
-              env,
-              timeoutMs: commandTimeoutMs,
-              needsScreenRecording,
-              agentId,
-              sessionKey,
-            },
-            timeoutMs: invokeTimeoutMs,
-            idempotencyKey: crypto.randomUUID(),
-          })) as { payload?: unknown };
-          return jsonResult(raw?.payload ?? {});
-        }
-        default:
-          throw new Error(`Unknown action: ${action}`);
+      } catch (err) {
+        const nodeLabel =
+          typeof params.node === "string" && params.node.trim() ? params.node.trim() : "auto";
+        const gatewayLabel =
+          gatewayOpts.gatewayUrl && gatewayOpts.gatewayUrl.trim()
+            ? gatewayOpts.gatewayUrl.trim()
+            : "default";
+        const agentLabel = agentId ?? "unknown";
+        const message = err instanceof Error ? err.message : String(err);
+        throw new Error(
+          `agent=${agentLabel} node=${nodeLabel} gateway=${gatewayLabel} action=${action}: ${message}`,
+        );
      }
    },
  };
@@ -121,6 +121,42 @@ describe("directive behavior", () => {
      expect(runEmbeddedPiAgent).not.toHaveBeenCalled();
    });
  });
+  it("includes catalog models when no allowlist is set", async () => {
+    await withTempHome(async (home) => {
+      vi.mocked(runEmbeddedPiAgent).mockReset();
+      vi.mocked(loadModelCatalog).mockResolvedValueOnce([
+        { id: "claude-opus-4-5", name: "Opus 4.5", provider: "anthropic" },
+        { id: "gpt-4.1-mini", name: "GPT-4.1 Mini", provider: "openai" },
+        { id: "grok-4", name: "Grok 4", provider: "xai" },
+      ]);
+      const storePath = path.join(home, "sessions.json");
+
+      const res = await getReplyFromConfig(
+        { Body: "/model list", From: "+1222", To: "+1222", CommandAuthorized: true },
+        {},
+        {
+          agents: {
+            defaults: {
+              model: {
+                primary: "anthropic/claude-opus-4-5",
+                fallbacks: ["openai/gpt-4.1-mini"],
+              },
+              imageModel: { primary: "minimax/MiniMax-M2.1" },
+              workspace: path.join(home, "clawd"),
+            },
+          },
+          session: { store: storePath },
+        },
+      );
+
+      const text = Array.isArray(res) ? res[0]?.text : res?.text;
+      expect(text).toContain("anthropic/claude-opus-4-5");
+      expect(text).toContain("openai/gpt-4.1-mini");
+      expect(text).toContain("minimax/MiniMax-M2.1");
+      expect(text).toContain("xai/grok-4");
+      expect(runEmbeddedPiAgent).not.toHaveBeenCalled();
+    });
+  });
  it("merges config allowlist models even when catalog is present", async () => {
    await withTempHome(async (home) => {
      vi.mocked(runEmbeddedPiAgent).mockReset();
@@ -299,6 +299,8 @@ export async function runAgentTurnWithFallback(params: {
                  const { text, skip } = normalizeStreamingText(payload);
                  const hasPayloadMedia = (payload.mediaUrls?.length ?? 0) > 0;
                  if (skip && !hasPayloadMedia) return;
+                  const currentMessageId =
+                    params.sessionCtx.MessageSidFull ?? params.sessionCtx.MessageSid;
                  const taggedPayload = applyReplyTagsToPayload(
                    {
                      text,
@@ -308,12 +310,12 @@ export async function runAgentTurnWithFallback(params: {
                      replyToTag: payload.replyToTag,
                      replyToCurrent: payload.replyToCurrent,
                    },
-                    params.sessionCtx.MessageSid,
+                    currentMessageId,
                  );
                  // Let through payloads with audioAsVoice flag even if empty (need to track it)
                  if (!isRenderablePayload(taggedPayload) && !payload.audioAsVoice) return;
                  const parsed = parseReplyDirectives(taggedPayload.text ?? "", {
-                    currentMessageId: params.sessionCtx.MessageSid,
+                    currentMessageId,
                    silentToken: SILENT_REPLY_TOKEN,
                  });
                  const cleaned = parsed.text || undefined;
@@ -1,6 +1,6 @@
 import type { NormalizedUsage } from "../../agents/usage.js";
 import { getChannelDock } from "../../channels/dock.js";
-import type { ChannelThreadingToolContext } from "../../channels/plugins/types.js";
+import type { ChannelId, ChannelThreadingToolContext } from "../../channels/plugins/types.js";
 import { normalizeChannelId } from "../../channels/registry.js";
 import type { ClawdbotConfig } from "../../config/config.js";
 import { isReasoningTagProvider } from "../../utils/provider-utils.js";
@@ -21,17 +21,25 @@ export function buildThreadingToolContext(params: {
 }): ChannelThreadingToolContext {
  const { sessionCtx, config, hasRepliedRef } = params;
  if (!config) return {};
-  const provider = normalizeChannelId(sessionCtx.Provider);
-  if (!provider) return {};
-  const dock = getChannelDock(provider);
-  if (!dock?.threading?.buildToolContext) return {};
+  const rawProvider = sessionCtx.Provider?.trim().toLowerCase();
+  if (!rawProvider) return {};
+  const provider = normalizeChannelId(rawProvider);
  // WhatsApp context isolation keys off conversation id, not the bot's own number.
  const threadingTo =
-    provider === "whatsapp"
+    rawProvider === "whatsapp"
      ? (sessionCtx.From ?? sessionCtx.To)
-      : provider === "imessage" && sessionCtx.ChatType === "direct"
+      : rawProvider === "imessage" && sessionCtx.ChatType === "direct"
        ? (sessionCtx.From ?? sessionCtx.To)
        : sessionCtx.To;
+  // Fallback for unrecognized/plugin channels (e.g., BlueBubbles before plugin registry init)
+  const dock = provider ? getChannelDock(provider) : undefined;
+  if (!dock?.threading?.buildToolContext) {
+    return {
+      currentChannelId: threadingTo?.trim() || undefined,
+      currentChannelProvider: provider ?? (rawProvider as ChannelId),
+      hasRepliedRef,
+    };
+  }
  const context =
    dock.threading.buildToolContext({
      cfg: config,
@@ -47,7 +55,7 @@ export function buildThreadingToolContext(params: {
    }) ?? {};
  return {
    ...context,
-    currentChannelProvider: provider,
+    currentChannelProvider: provider!, // guaranteed non-null since dock exists
  };
 }

@@ -377,7 +377,7 @@ export async function runReplyAgent(params: {
      directlySentBlockKeys,
      replyToMode,
      replyToChannel,
-      currentMessageId: sessionCtx.MessageSid,
+      currentMessageId: sessionCtx.MessageSidFull ?? sessionCtx.MessageSid,
      messageProvider: followupRun.run.messageProvider,
      messagingToolSentTexts: runResult.messagingToolSentTexts,
      messagingToolSentTargets: runResult.messagingToolSentTargets,
@@ -36,6 +36,56 @@ function buildModelPickerCatalog(params: {
    defaultModel: params.defaultModel,
  });

+  const buildConfiguredCatalog = (): ModelPickerCatalogEntry[] => {
+    const out: ModelPickerCatalogEntry[] = [];
+    const keys = new Set<string>();
+
+    const pushRef = (ref: { provider: string; model: string }, name?: string) => {
+      const provider = normalizeProviderId(ref.provider);
+      const id = String(ref.model ?? "").trim();
+      if (!provider || !id) return;
+      const key = modelKey(provider, id);
+      if (keys.has(key)) return;
+      keys.add(key);
+      out.push({ provider, id, name: name ?? id });
+    };
+
+    const pushRaw = (raw?: string) => {
+      const value = String(raw ?? "").trim();
+      if (!value) return;
+      const resolved = resolveModelRefFromString({
+        raw: value,
+        defaultProvider: params.defaultProvider,
+        aliasIndex: params.aliasIndex,
+      });
+      if (!resolved) return;
+      pushRef(resolved.ref);
+    };
+
+    pushRef(resolvedDefault);
+
+    const modelConfig = params.cfg.agents?.defaults?.model;
+    const modelFallbacks =
+      modelConfig && typeof modelConfig === "object" ? (modelConfig.fallbacks ?? []) : [];
+    for (const fallback of modelFallbacks) {
+      pushRaw(String(fallback ?? ""));
+    }
+
+    const imageConfig = params.cfg.agents?.defaults?.imageModel;
+    if (imageConfig && typeof imageConfig === "object") {
+      pushRaw(imageConfig.primary);
+      for (const fallback of imageConfig.fallbacks ?? []) {
+        pushRaw(String(fallback ?? ""));
+      }
+    }
+
+    for (const raw of Object.keys(params.cfg.agents?.defaults?.models ?? {})) {
+      pushRaw(raw);
+    }
+
+    return out;
+  };
+
  const keys = new Set<string>();
  const out: ModelPickerCatalogEntry[] = [];

@@ -49,6 +99,21 @@ function buildModelPickerCatalog(params: {
    out.push({ provider, id, name: entry.name });
  };

+  const hasAllowlist = Object.keys(params.cfg.agents?.defaults?.models ?? {}).length > 0;
+  if (!hasAllowlist) {
+    for (const entry of params.allowedModelCatalog) {
+      push({
+        provider: entry.provider,
+        id: entry.id ?? "",
+        name: entry.name,
+      });
+    }
+    for (const entry of buildConfiguredCatalog()) {
+      push(entry);
+    }
+    return out;
+  }
+
  // Prefer catalog entries (when available), but always merge in config-only
  // allowlist entries. This keeps custom providers/models visible in /model.
  for (const entry of params.allowedModelCatalog) {
@@ -350,7 +350,7 @@ export async function runPreparedReply(
  const authProfileIdSource = sessionEntry?.authProfileOverrideSource;
  const followupRun = {
    prompt: queuedBody,
-    messageId: sessionCtx.MessageSid,
+    messageId: sessionCtx.MessageSidFull ?? sessionCtx.MessageSid,
    summaryLine: baseBodyTrimmedRaw,
    enqueuedAt: Date.now(),
    // Originating channel for reply routing.
@@ -140,7 +140,7 @@ describe("createTypingSignaler", () => {
    expect(typing.startTypingOnText).not.toHaveBeenCalled();
  });

-  it("does not start typing on tool start before text", async () => {
+  it("starts typing on tool start before text", async () => {
    const typing = createMockTypingController();
    const signaler = createTypingSignaler({
      typing,
@@ -150,8 +150,9 @@ describe("createTypingSignaler", () => {

    await signaler.signalToolStart();

-    expect(typing.startTypingLoop).not.toHaveBeenCalled();
-    expect(typing.refreshTypingTtl).not.toHaveBeenCalled();
+    expect(typing.startTypingLoop).toHaveBeenCalled();
+    expect(typing.refreshTypingTtl).toHaveBeenCalled();
+    expect(typing.startTypingOnText).not.toHaveBeenCalled();
  });

  it("refreshes ttl on tool start when active after text", async () => {
@@ -95,13 +95,13 @@ export function createTypingSignaler(params: {

  const signalToolStart = async () => {
    if (disabled) return;
-    if (!hasRenderableText) return;
+    // Start typing as soon as tools begin executing, even before the first text delta.
    if (!typing.isActive()) {
      await typing.startTypingLoop();
      typing.refreshTypingTtl();
      return;
    }
-    // Keep typing indicator alive during tool execution without changing mode semantics.
+    // Keep typing indicator alive during tool execution.
    typing.refreshTypingTtl();
  };

@@ -74,6 +74,25 @@ describe("buildStatusMessage", () => {
    expect(normalized).toContain("Queue: collect");
  });

+  it("uses per-agent sandbox config when config and session key are provided", () => {
+    const text = buildStatusMessage({
+      config: {
+        agents: {
+          list: [
+            { id: "main", default: true },
+            { id: "discord", sandbox: { mode: "all" } },
+          ],
+        },
+      } as ClawdbotConfig,
+      agent: {},
+      sessionKey: "agent:discord:discord:channel:1456350065223270435",
+      sessionScope: "per-sender",
+      queue: { mode: "collect", depth: 0 },
+    });
+
+    expect(normalizeTestText(text)).toContain("Runtime: docker/all");
+  });
+
  it("shows verbose/elevated labels only when enabled", () => {
    const text = buildStatusMessage({
      agent: { model: "anthropic/claude-opus-4-5" },
@@ -59,6 +59,42 @@ type StatusArgs = {
  now?: number;
 };

+function resolveRuntimeLabel(
+  args: Pick<StatusArgs, "config" | "agent" | "sessionKey" | "sessionScope">,
+): string {
+  const sessionKey = args.sessionKey?.trim();
+  if (args.config && sessionKey) {
+    const runtimeStatus = resolveSandboxRuntimeStatus({
+      cfg: args.config,
+      sessionKey,
+    });
+    const sandboxMode = runtimeStatus.mode ?? "off";
+    if (sandboxMode === "off") return "direct";
+    const runtime = runtimeStatus.sandboxed ? "docker" : sessionKey ? "direct" : "unknown";
+    return `${runtime}/${sandboxMode}`;
+  }
+
+  const sandboxMode = args.agent?.sandbox?.mode ?? "off";
+  if (sandboxMode === "off") return "direct";
+  const sandboxed = (() => {
+    if (!sessionKey) return false;
+    if (sandboxMode === "all") return true;
+    if (args.config) {
+      return resolveSandboxRuntimeStatus({
+        cfg: args.config,
+        sessionKey,
+      }).sandboxed;
+    }
+    const sessionScope = args.sessionScope ?? "per-sender";
+    const mainKey = resolveMainSessionKey({
+      session: { scope: sessionScope },
+    });
+    return sessionKey !== mainKey.trim();
+  })();
+  const runtime = sandboxed ? "docker" : sessionKey ? "direct" : "unknown";
+  return `${runtime}/${sandboxMode}`;
+}
+
 const formatTokens = (total: number | null | undefined, contextTokens: number | null) => {
  const ctx = contextTokens ?? null;
  if (total == null) {
@@ -258,30 +294,7 @@ export function buildStatusMessage(args: StatusArgs): string {
    args.agent?.elevatedDefault ??
    "on";

-  const runtime = (() => {
-    const sandboxMode = args.agent?.sandbox?.mode ?? "off";
-    if (sandboxMode === "off") return { label: "direct" };
-    const sessionKey = args.sessionKey?.trim();
-    const sandboxed = (() => {
-      if (!sessionKey) return false;
-      if (sandboxMode === "all") return true;
-      if (args.config) {
-        return resolveSandboxRuntimeStatus({
-          cfg: args.config,
-          sessionKey,
-        }).sandboxed;
-      }
-      const sessionScope = args.sessionScope ?? "per-sender";
-      const mainKey = resolveMainSessionKey({
-        session: { scope: sessionScope },
-      });
-      return sessionKey !== mainKey.trim();
-    })();
-    const runtime = sandboxed ? "docker" : sessionKey ? "direct" : "unknown";
-    return {
-      label: `${runtime}/${sandboxMode}`,
-    };
-  })();
+  const runtime = { label: resolveRuntimeLabel(args) };

  const updatedAt = entry?.updatedAt;
  const sessionLine = [
@@ -38,10 +38,14 @@ export type MsgContext = {
  AccountId?: string;
  ParentSessionKey?: string;
  MessageSid?: string;
+  /** Provider-specific full message id when MessageSid is a shortened alias. */
+  MessageSidFull?: string;
  MessageSids?: string[];
  MessageSidFirst?: string;
  MessageSidLast?: string;
  ReplyToId?: string;
+  /** Provider-specific full reply-to id when ReplyToId is a shortened alias. */
+  ReplyToIdFull?: string;
  ReplyToBody?: string;
  ReplyToSender?: string;
  ForwardedFrom?: string;
@@ -212,6 +212,7 @@ export type ChannelThreadingContext = {
  Channel?: string;
  To?: string;
  ReplyToId?: string;
+  ReplyToIdFull?: string;
  ThreadLabel?: string;
  MessageThreadId?: string | number;
 };
@@ -23,6 +23,14 @@ const defaultRuntime = {
  },
 };

+const localSnapshot = {
+  path: "/tmp/local-exec-approvals.json",
+  exists: true,
+  raw: "{}",
+  hash: "hash-local",
+  file: { version: 1, agents: {} },
+};
+
 vi.mock("./gateway-rpc.js", () => ({
  callGatewayFromCli: (method: string, opts: unknown, params?: unknown) =>
    callGatewayFromCli(method, opts, params),
@@ -40,8 +48,19 @@ vi.mock("../runtime.js", () => ({
  defaultRuntime,
 }));

+vi.mock("../infra/exec-approvals.js", async () => {
+  const actual = await vi.importActual<typeof import("../infra/exec-approvals.js")>(
+    "../infra/exec-approvals.js",
+  );
+  return {
+    ...actual,
+    readExecApprovalsSnapshot: () => localSnapshot,
+    saveExecApprovals: vi.fn(),
+  };
+});
+
 describe("exec approvals CLI", () => {
-  it("loads gateway approvals by default", async () => {
+  it("loads local approvals by default", async () => {
    runtimeLogs.length = 0;
    runtimeErrors.length = 0;
    callGatewayFromCli.mockClear();
@@ -53,6 +72,22 @@ describe("exec approvals CLI", () => {

    await program.parseAsync(["approvals", "get"], { from: "user" });

+    expect(callGatewayFromCli).not.toHaveBeenCalled();
+    expect(runtimeErrors).toHaveLength(0);
+  });
+
+  it("loads gateway approvals when --gateway is set", async () => {
+    runtimeLogs.length = 0;
+    runtimeErrors.length = 0;
+    callGatewayFromCli.mockClear();
+
+    const { registerExecApprovalsCli } = await import("./exec-approvals-cli.js");
+    const program = new Command();
+    program.exitOverride();
+    registerExecApprovalsCli(program);
+
+    await program.parseAsync(["approvals", "get", "--gateway"], { from: "user" });
+
    expect(callGatewayFromCli).toHaveBeenCalledWith("exec.approvals.get", expect.anything(), {});
    expect(runtimeErrors).toHaveLength(0);
  });
@@ -74,4 +109,34 @@ describe("exec approvals CLI", () => {
    });
    expect(runtimeErrors).toHaveLength(0);
  });
+
+  it("defaults allowlist add to wildcard agent", async () => {
+    runtimeLogs.length = 0;
+    runtimeErrors.length = 0;
+    callGatewayFromCli.mockClear();
+
+    const execApprovals = await import("../infra/exec-approvals.js");
+    const saveExecApprovals = vi.mocked(execApprovals.saveExecApprovals);
+    saveExecApprovals.mockClear();
+
+    const { registerExecApprovalsCli } = await import("./exec-approvals-cli.js");
+    const program = new Command();
+    program.exitOverride();
+    registerExecApprovalsCli(program);
+
+    await program.parseAsync(["approvals", "allowlist", "add", "/usr/bin/uname"], { from: "user" });
+
+    expect(callGatewayFromCli).not.toHaveBeenCalledWith(
+      "exec.approvals.set",
+      expect.anything(),
+      {},
+    );
+    expect(saveExecApprovals).toHaveBeenCalledWith(
+      expect.objectContaining({
+        agents: expect.objectContaining({
+          "*": expect.anything(),
+        }),
+      }),
+    );
+  });
 });
@@ -2,12 +2,18 @@ import fs from "node:fs/promises";
 import JSON5 from "json5";
 import type { Command } from "commander";

-import type { ExecApprovalsAgent, ExecApprovalsFile } from "../infra/exec-approvals.js";
+import {
+  readExecApprovalsSnapshot,
+  saveExecApprovals,
+  type ExecApprovalsAgent,
+  type ExecApprovalsFile,
+} from "../infra/exec-approvals.js";
 import { defaultRuntime } from "../runtime.js";
 import { formatDocsLink } from "../terminal/links.js";
 import { isRich, theme } from "../terminal/theme.js";
 import { renderTable } from "../terminal/table.js";
 import { callGatewayFromCli } from "./gateway-rpc.js";
+import { describeUnknownError } from "./gateway-cli/shared.js";
 import { nodesCallOpts, resolveNodeId } from "./nodes-cli/rpc.js";
 import type { NodesRpcOpts } from "./nodes-cli/types.js";

@@ -20,6 +26,7 @@ type ExecApprovalsSnapshot = {

 type ExecApprovalsCliOpts = NodesRpcOpts & {
  node?: string;
+  gateway?: boolean;
  file?: string;
  stdin?: boolean;
  agent?: string;
@@ -45,6 +52,7 @@ async function readStdin(): Promise<string> {
 }

 async function resolveTargetNodeId(opts: ExecApprovalsCliOpts): Promise<string | null> {
+  if (opts.gateway) return null;
  const raw = opts.node?.trim() ?? "";
  if (!raw) return null;
  return await resolveNodeId(opts as NodesRpcOpts, raw);
@@ -60,6 +68,120 @@ async function loadSnapshot(
  return snapshot;
 }

+function loadSnapshotLocal(): ExecApprovalsSnapshot {
+  const snapshot = readExecApprovalsSnapshot();
+  return {
+    path: snapshot.path,
+    exists: snapshot.exists,
+    hash: snapshot.hash,
+    file: snapshot.file,
+  };
+}
+
+function saveSnapshotLocal(file: ExecApprovalsFile): ExecApprovalsSnapshot {
+  saveExecApprovals(file);
+  return loadSnapshotLocal();
+}
+
+async function loadSnapshotTarget(opts: ExecApprovalsCliOpts): Promise<{
+  snapshot: ExecApprovalsSnapshot;
+  nodeId: string | null;
+  source: "gateway" | "node" | "local";
+}> {
+  if (!opts.gateway && !opts.node) {
+    return { snapshot: loadSnapshotLocal(), nodeId: null, source: "local" };
+  }
+  const nodeId = await resolveTargetNodeId(opts);
+  const snapshot = await loadSnapshot(opts, nodeId);
+  return { snapshot, nodeId, source: nodeId ? "node" : "gateway" };
+}
+
+function formatCliError(err: unknown): string {
+  const msg = describeUnknownError(err);
+  return msg.includes("\n") ? msg.split("\n")[0] : msg;
+}
+
+function renderApprovalsSnapshot(snapshot: ExecApprovalsSnapshot, targetLabel: string) {
+  const rich = isRich();
+  const heading = (text: string) => (rich ? theme.heading(text) : text);
+  const muted = (text: string) => (rich ? theme.muted(text) : text);
+  const tableWidth = Math.max(60, (process.stdout.columns ?? 120) - 1);
+
+  const file = snapshot.file ?? { version: 1 };
+  const defaults = file.defaults ?? {};
+  const defaultsParts = [
+    defaults.security ? `security=${defaults.security}` : null,
+    defaults.ask ? `ask=${defaults.ask}` : null,
+    defaults.askFallback ? `askFallback=${defaults.askFallback}` : null,
+    typeof defaults.autoAllowSkills === "boolean"
+      ? `autoAllowSkills=${defaults.autoAllowSkills ? "on" : "off"}`
+      : null,
+  ].filter(Boolean) as string[];
+  const agents = file.agents ?? {};
+  const allowlistRows: Array<{ Target: string; Agent: string; Pattern: string; LastUsed: string }> =
+    [];
+  const now = Date.now();
+  for (const [agentId, agent] of Object.entries(agents)) {
+    const allowlist = Array.isArray(agent.allowlist) ? agent.allowlist : [];
+    for (const entry of allowlist) {
+      const pattern = entry?.pattern?.trim() ?? "";
+      if (!pattern) continue;
+      const lastUsedAt = typeof entry.lastUsedAt === "number" ? entry.lastUsedAt : null;
+      allowlistRows.push({
+        Target: targetLabel,
+        Agent: agentId,
+        Pattern: pattern,
+        LastUsed: lastUsedAt ? `${formatAge(Math.max(0, now - lastUsedAt))} ago` : muted("unknown"),
+      });
+    }
+  }
+
+  const summaryRows = [
+    { Field: "Target", Value: targetLabel },
+    { Field: "Path", Value: snapshot.path },
+    { Field: "Exists", Value: snapshot.exists ? "yes" : "no" },
+    { Field: "Hash", Value: snapshot.hash },
+    { Field: "Version", Value: String(file.version ?? 1) },
+    { Field: "Socket", Value: file.socket?.path ?? "default" },
+    { Field: "Defaults", Value: defaultsParts.length > 0 ? defaultsParts.join(", ") : "none" },
+    { Field: "Agents", Value: String(Object.keys(agents).length) },
+    { Field: "Allowlist", Value: String(allowlistRows.length) },
+  ];
+
+  defaultRuntime.log(heading("Approvals"));
+  defaultRuntime.log(
+    renderTable({
+      width: tableWidth,
+      columns: [
+        { key: "Field", header: "Field", minWidth: 8 },
+        { key: "Value", header: "Value", minWidth: 24, flex: true },
+      ],
+      rows: summaryRows,
+    }).trimEnd(),
+  );
+
+  if (allowlistRows.length === 0) {
+    defaultRuntime.log("");
+    defaultRuntime.log(muted("No allowlist entries."));
+    return;
+  }
+
+  defaultRuntime.log("");
+  defaultRuntime.log(heading("Allowlist"));
+  defaultRuntime.log(
+    renderTable({
+      width: tableWidth,
+      columns: [
+        { key: "Target", header: "Target", minWidth: 10 },
+        { key: "Agent", header: "Agent", minWidth: 8 },
+        { key: "Pattern", header: "Pattern", minWidth: 20, flex: true },
+        { key: "LastUsed", header: "Last Used", minWidth: 10 },
+      ],
+      rows: allowlistRows,
+    }).trimEnd(),
+  );
+}
+
 async function saveSnapshot(
  opts: ExecApprovalsCliOpts,
  nodeId: string | null,
@@ -74,7 +196,7 @@ async function saveSnapshot(

 function resolveAgentKey(value?: string | null): string {
  const trimmed = value?.trim() ?? "";
-  return trimmed ? trimmed : "default";
+  return trimmed ? trimmed : "*";
 }

 function normalizeAllowlistEntry(entry: { pattern?: string } | null): string | null {
@@ -117,131 +239,83 @@ export function registerExecApprovalsCli(program: Command) {
  const getCmd = approvals
    .command("get")
    .description("Fetch exec approvals snapshot")
-    .option("--node <node>", "Target node id/name/IP (defaults to gateway)")
+    .option("--node <node>", "Target node id/name/IP")
+    .option("--gateway", "Force gateway approvals", false)
    .action(async (opts: ExecApprovalsCliOpts) => {
-      const nodeId = await resolveTargetNodeId(opts);
-      const snapshot = await loadSnapshot(opts, nodeId);
-      if (opts.json) {
-        defaultRuntime.log(JSON.stringify(snapshot));
-        return;
-      }
-
-      const rich = isRich();
-      const heading = (text: string) => (rich ? theme.heading(text) : text);
-      const muted = (text: string) => (rich ? theme.muted(text) : text);
-      const tableWidth = Math.max(60, (process.stdout.columns ?? 120) - 1);
-
-      const file = snapshot.file ?? { version: 1 };
-      const defaults = file.defaults ?? {};
-      const defaultsParts = [
-        defaults.security ? `security=${defaults.security}` : null,
-        defaults.ask ? `ask=${defaults.ask}` : null,
-        defaults.askFallback ? `askFallback=${defaults.askFallback}` : null,
-        typeof defaults.autoAllowSkills === "boolean"
-          ? `autoAllowSkills=${defaults.autoAllowSkills ? "on" : "off"}`
-          : null,
-      ].filter(Boolean) as string[];
-      const agents = file.agents ?? {};
-      const allowlistRows: Array<{ Agent: string; Pattern: string; LastUsed: string }> = [];
-      const now = Date.now();
-      for (const [agentId, agent] of Object.entries(agents)) {
-        const allowlist = Array.isArray(agent.allowlist) ? agent.allowlist : [];
-        for (const entry of allowlist) {
-          const pattern = entry?.pattern?.trim() ?? "";
-          if (!pattern) continue;
-          const lastUsedAt = typeof entry.lastUsedAt === "number" ? entry.lastUsedAt : null;
-          allowlistRows.push({
-            Agent: agentId,
-            Pattern: pattern,
-            LastUsed: lastUsedAt
-              ? `${formatAge(Math.max(0, now - lastUsedAt))} ago`
-              : muted("unknown"),
-          });
+      try {
+        const { snapshot, nodeId, source } = await loadSnapshotTarget(opts);
+        if (opts.json) {
+          defaultRuntime.log(JSON.stringify(snapshot));
+          return;
        }
+
+        const muted = (text: string) => (isRich() ? theme.muted(text) : text);
+        if (source === "local") {
+          defaultRuntime.log(muted("Showing local approvals."));
+          defaultRuntime.log("");
+        }
+        const targetLabel = source === "local" ? "local" : nodeId ? `node:${nodeId}` : "gateway";
+        renderApprovalsSnapshot(snapshot, targetLabel);
+      } catch (err) {
+        defaultRuntime.error(formatCliError(err));
+        defaultRuntime.exit(1);
      }
-
-      const summaryRows = [
-        { Field: "Path", Value: snapshot.path },
-        { Field: "Exists", Value: snapshot.exists ? "yes" : "no" },
-        { Field: "Hash", Value: snapshot.hash },
-        { Field: "Version", Value: String(file.version ?? 1) },
-        { Field: "Socket", Value: file.socket?.path ?? "default" },
-        { Field: "Defaults", Value: defaultsParts.length > 0 ? defaultsParts.join(", ") : "none" },
-        { Field: "Agents", Value: String(Object.keys(agents).length) },
-        { Field: "Allowlist", Value: String(allowlistRows.length) },
-      ];
-
-      defaultRuntime.log(heading("Approvals"));
-      defaultRuntime.log(
-        renderTable({
-          width: tableWidth,
-          columns: [
-            { key: "Field", header: "Field", minWidth: 8 },
-            { key: "Value", header: "Value", minWidth: 24, flex: true },
-          ],
-          rows: summaryRows,
-        }).trimEnd(),
-      );
-
-      if (allowlistRows.length === 0) {
-        defaultRuntime.log("");
-        defaultRuntime.log(muted("No allowlist entries."));
-        return;
-      }
-
-      defaultRuntime.log("");
-      defaultRuntime.log(heading("Allowlist"));
-      defaultRuntime.log(
-        renderTable({
-          width: tableWidth,
-          columns: [
-            { key: "Agent", header: "Agent", minWidth: 8 },
-            { key: "Pattern", header: "Pattern", minWidth: 20, flex: true },
-            { key: "LastUsed", header: "Last Used", minWidth: 10 },
-          ],
-          rows: allowlistRows,
-        }).trimEnd(),
-      );
    });
  nodesCallOpts(getCmd);

  const setCmd = approvals
    .command("set")
    .description("Replace exec approvals with a JSON file")
-    .option("--node <node>", "Target node id/name/IP (defaults to gateway)")
+    .option("--node <node>", "Target node id/name/IP")
+    .option("--gateway", "Force gateway approvals", false)
    .option("--file <path>", "Path to JSON file to upload")
    .option("--stdin", "Read JSON from stdin", false)
    .action(async (opts: ExecApprovalsCliOpts) => {
-      if (!opts.file && !opts.stdin) {
-        defaultRuntime.error("Provide --file or --stdin.");
-        defaultRuntime.exit(1);
-        return;
-      }
-      if (opts.file && opts.stdin) {
-        defaultRuntime.error("Use either --file or --stdin (not both).");
-        defaultRuntime.exit(1);
-        return;
-      }
-      const nodeId = await resolveTargetNodeId(opts);
-      const snapshot = await loadSnapshot(opts, nodeId);
-      if (!snapshot.hash) {
-        defaultRuntime.error("Exec approvals hash missing; reload and retry.");
-        defaultRuntime.exit(1);
-        return;
-      }
-      const raw = opts.stdin ? await readStdin() : await fs.readFile(String(opts.file), "utf8");
-      let file: ExecApprovalsFile;
      try {
-        file = JSON5.parse(raw) as ExecApprovalsFile;
+        if (!opts.file && !opts.stdin) {
+          defaultRuntime.error("Provide --file or --stdin.");
+          defaultRuntime.exit(1);
+          return;
+        }
+        if (opts.file && opts.stdin) {
+          defaultRuntime.error("Use either --file or --stdin (not both).");
+          defaultRuntime.exit(1);
+          return;
+        }
+        const { snapshot, nodeId, source } = await loadSnapshotTarget(opts);
+        if (source === "local") {
+          defaultRuntime.log(theme.muted("Writing local approvals."));
+        }
+        const targetLabel = source === "local" ? "local" : nodeId ? `node:${nodeId}` : "gateway";
+        if (!snapshot.hash) {
+          defaultRuntime.error("Exec approvals hash missing; reload and retry.");
+          defaultRuntime.exit(1);
+          return;
+        }
+        const raw = opts.stdin ? await readStdin() : await fs.readFile(String(opts.file), "utf8");
+        let file: ExecApprovalsFile;
+        try {
+          file = JSON5.parse(raw) as ExecApprovalsFile;
+        } catch (err) {
+          defaultRuntime.error(`Failed to parse approvals JSON: ${String(err)}`);
+          defaultRuntime.exit(1);
+          return;
+        }
+        file.version = 1;
+        const next =
+          source === "local"
+            ? saveSnapshotLocal(file)
+            : await saveSnapshot(opts, nodeId, file, snapshot.hash);
+        if (opts.json) {
+          defaultRuntime.log(JSON.stringify(next));
+          return;
+        }
+        defaultRuntime.log(theme.muted(`Target: ${targetLabel}`));
+        renderApprovalsSnapshot(next, targetLabel);
      } catch (err) {
-        defaultRuntime.error(`Failed to parse approvals JSON: ${String(err)}`);
+        defaultRuntime.error(formatCliError(err));
        defaultRuntime.exit(1);
-        return;
      }
-      file.version = 1;
-      const next = await saveSnapshot(opts, nodeId, file, snapshot.hash);
-      const payload = opts.json ? JSON.stringify(next) : JSON.stringify(next, null, 2);
-      defaultRuntime.log(payload);
    });
  nodesCallOpts(setCmd);

@@ -253,10 +327,13 @@ export function registerExecApprovalsCli(program: Command) {
      () =>
        `\n${theme.heading("Examples:")}\n${formatExample(
          'clawdbot approvals allowlist add "~/Projects/**/bin/rg"',
-          "Allowlist a local binary pattern for the default agent.",
+          "Allowlist a local binary pattern for the main agent.",
        )}\n${formatExample(
          'clawdbot approvals allowlist add --agent main --node <id|name|ip> "/usr/bin/uptime"',
          "Allowlist on a specific node/agent.",
+        )}\n${formatExample(
+          'clawdbot approvals allowlist add --agent "*" "/usr/bin/uname"',
+          "Allowlist for all agents (wildcard).",
        )}\n${formatExample(
          'clawdbot approvals allowlist remove "~/Projects/**/bin/rg"',
          "Remove an allowlist pattern.",
@@ -266,86 +343,118 @@ export function registerExecApprovalsCli(program: Command) {
  const allowlistAdd = allowlist
    .command("add <pattern>")
    .description("Add a glob pattern to an allowlist")
-    .option("--node <node>", "Target node id/name/IP (defaults to gateway)")
-    .option("--agent <id>", 'Agent id (defaults to "default")')
+    .option("--node <node>", "Target node id/name/IP")
+    .option("--gateway", "Force gateway approvals", false)
+    .option("--agent <id>", 'Agent id (defaults to "*")')
    .action(async (pattern: string, opts: ExecApprovalsCliOpts) => {
-      const trimmed = pattern.trim();
-      if (!trimmed) {
-        defaultRuntime.error("Pattern required.");
+      try {
+        const trimmed = pattern.trim();
+        if (!trimmed) {
+          defaultRuntime.error("Pattern required.");
+          defaultRuntime.exit(1);
+          return;
+        }
+        const { snapshot, nodeId, source } = await loadSnapshotTarget(opts);
+        if (source === "local") {
+          defaultRuntime.log(theme.muted("Writing local approvals."));
+        }
+        const targetLabel = source === "local" ? "local" : nodeId ? `node:${nodeId}` : "gateway";
+        if (!snapshot.hash) {
+          defaultRuntime.error("Exec approvals hash missing; reload and retry.");
+          defaultRuntime.exit(1);
+          return;
+        }
+        const file = snapshot.file ?? { version: 1 };
+        file.version = 1;
+        const agentKey = resolveAgentKey(opts.agent);
+        const agent = ensureAgent(file, agentKey);
+        const allowlistEntries = Array.isArray(agent.allowlist) ? agent.allowlist : [];
+        if (allowlistEntries.some((entry) => normalizeAllowlistEntry(entry) === trimmed)) {
+          defaultRuntime.log("Already allowlisted.");
+          return;
+        }
+        allowlistEntries.push({ pattern: trimmed, lastUsedAt: Date.now() });
+        agent.allowlist = allowlistEntries;
+        file.agents = { ...file.agents, [agentKey]: agent };
+        const next =
+          source === "local"
+            ? saveSnapshotLocal(file)
+            : await saveSnapshot(opts, nodeId, file, snapshot.hash);
+        if (opts.json) {
+          defaultRuntime.log(JSON.stringify(next));
+          return;
+        }
+        defaultRuntime.log(theme.muted(`Target: ${targetLabel}`));
+        renderApprovalsSnapshot(next, targetLabel);
+      } catch (err) {
+        defaultRuntime.error(formatCliError(err));
        defaultRuntime.exit(1);
-        return;
      }
-      const nodeId = await resolveTargetNodeId(opts);
-      const snapshot = await loadSnapshot(opts, nodeId);
-      if (!snapshot.hash) {
-        defaultRuntime.error("Exec approvals hash missing; reload and retry.");
-        defaultRuntime.exit(1);
-        return;
-      }
-      const file = snapshot.file ?? { version: 1 };
-      file.version = 1;
-      const agentKey = resolveAgentKey(opts.agent);
-      const agent = ensureAgent(file, agentKey);
-      const allowlistEntries = Array.isArray(agent.allowlist) ? agent.allowlist : [];
-      if (allowlistEntries.some((entry) => normalizeAllowlistEntry(entry) === trimmed)) {
-        defaultRuntime.log("Already allowlisted.");
-        return;
-      }
-      allowlistEntries.push({ pattern: trimmed, lastUsedAt: Date.now() });
-      agent.allowlist = allowlistEntries;
-      file.agents = { ...file.agents, [agentKey]: agent };
-      const next = await saveSnapshot(opts, nodeId, file, snapshot.hash);
-      const payload = opts.json ? JSON.stringify(next) : JSON.stringify(next, null, 2);
-      defaultRuntime.log(payload);
    });
  nodesCallOpts(allowlistAdd);

  const allowlistRemove = allowlist
    .command("remove <pattern>")
    .description("Remove a glob pattern from an allowlist")
-    .option("--node <node>", "Target node id/name/IP (defaults to gateway)")
-    .option("--agent <id>", 'Agent id (defaults to "default")')
+    .option("--node <node>", "Target node id/name/IP")
+    .option("--gateway", "Force gateway approvals", false)
+    .option("--agent <id>", 'Agent id (defaults to "*")')
    .action(async (pattern: string, opts: ExecApprovalsCliOpts) => {
-      const trimmed = pattern.trim();
-      if (!trimmed) {
-        defaultRuntime.error("Pattern required.");
+      try {
+        const trimmed = pattern.trim();
+        if (!trimmed) {
+          defaultRuntime.error("Pattern required.");
+          defaultRuntime.exit(1);
+          return;
+        }
+        const { snapshot, nodeId, source } = await loadSnapshotTarget(opts);
+        if (source === "local") {
+          defaultRuntime.log(theme.muted("Writing local approvals."));
+        }
+        const targetLabel = source === "local" ? "local" : nodeId ? `node:${nodeId}` : "gateway";
+        if (!snapshot.hash) {
+          defaultRuntime.error("Exec approvals hash missing; reload and retry.");
+          defaultRuntime.exit(1);
+          return;
+        }
+        const file = snapshot.file ?? { version: 1 };
+        file.version = 1;
+        const agentKey = resolveAgentKey(opts.agent);
+        const agent = ensureAgent(file, agentKey);
+        const allowlistEntries = Array.isArray(agent.allowlist) ? agent.allowlist : [];
+        const nextEntries = allowlistEntries.filter(
+          (entry) => normalizeAllowlistEntry(entry) !== trimmed,
+        );
+        if (nextEntries.length === allowlistEntries.length) {
+          defaultRuntime.log("Pattern not found.");
+          return;
+        }
+        if (nextEntries.length === 0) {
+          delete agent.allowlist;
+        } else {
+          agent.allowlist = nextEntries;
+        }
+        if (isEmptyAgent(agent)) {
+          const agents = { ...file.agents };
+          delete agents[agentKey];
+          file.agents = Object.keys(agents).length > 0 ? agents : undefined;
+        } else {
+          file.agents = { ...file.agents, [agentKey]: agent };
+        }
+        const next =
+          source === "local"
+            ? saveSnapshotLocal(file)
+            : await saveSnapshot(opts, nodeId, file, snapshot.hash);
+        if (opts.json) {
+          defaultRuntime.log(JSON.stringify(next));
+          return;
+        }
+        defaultRuntime.log(theme.muted(`Target: ${targetLabel}`));
+        renderApprovalsSnapshot(next, targetLabel);
+      } catch (err) {
+        defaultRuntime.error(formatCliError(err));
        defaultRuntime.exit(1);
-        return;
      }
-      const nodeId = await resolveTargetNodeId(opts);
-      const snapshot = await loadSnapshot(opts, nodeId);
-      if (!snapshot.hash) {
-        defaultRuntime.error("Exec approvals hash missing; reload and retry.");
-        defaultRuntime.exit(1);
-        return;
-      }
-      const file = snapshot.file ?? { version: 1 };
-      file.version = 1;
-      const agentKey = resolveAgentKey(opts.agent);
-      const agent = ensureAgent(file, agentKey);
-      const allowlistEntries = Array.isArray(agent.allowlist) ? agent.allowlist : [];
-      const nextEntries = allowlistEntries.filter(
-        (entry) => normalizeAllowlistEntry(entry) !== trimmed,
-      );
-      if (nextEntries.length === allowlistEntries.length) {
-        defaultRuntime.log("Pattern not found.");
-        return;
-      }
-      if (nextEntries.length === 0) {
-        delete agent.allowlist;
-      } else {
-        agent.allowlist = nextEntries;
-      }
-      if (isEmptyAgent(agent)) {
-        const agents = { ...file.agents };
-        delete agents[agentKey];
-        file.agents = Object.keys(agents).length > 0 ? agents : undefined;
-      } else {
-        file.agents = { ...file.agents, [agentKey]: agent };
-      }
-      const next = await saveSnapshot(opts, nodeId, file, snapshot.hash);
-      const payload = opts.json ? JSON.stringify(next) : JSON.stringify(next, null, 2);
-      defaultRuntime.log(payload);
    });
  nodesCallOpts(allowlistRemove);
 }
@@ -47,10 +47,7 @@ type NodeDaemonStatusOptions = {
 };

 function renderNodeServiceStartHints(): string[] {
-  const base = [
-    formatCliCommand("clawdbot node service install"),
-    formatCliCommand("clawdbot node start"),
-  ];
+  const base = [formatCliCommand("clawdbot node install"), formatCliCommand("clawdbot node start")];
  switch (process.platform) {
    case "darwin":
      return [
@@ -172,9 +169,7 @@ export async function runNodeDaemonInstall(opts: NodeDaemonInstallOptions) {
    });
    if (!json) {
      defaultRuntime.log(`Node service already ${service.loadedText}.`);
-      defaultRuntime.log(
-        `Reinstall with: ${formatCliCommand("clawdbot node service install --force")}`,
-      );
+      defaultRuntime.log(`Reinstall with: ${formatCliCommand("clawdbot node install --force")}`);
    }
    return;
  }
@@ -6,7 +6,6 @@ import { runNodeHost } from "../../node-host/runner.js";
 import {
  runNodeDaemonInstall,
  runNodeDaemonRestart,
-  runNodeDaemonStart,
  runNodeDaemonStatus,
  runNodeDaemonStop,
  runNodeDaemonUninstall,
@@ -28,13 +27,13 @@ export function registerNodeCli(program: Command) {
    );

  node
-    .command("start")
-    .description("Start the headless node host (foreground)")
+    .command("run")
+    .description("Run the headless node host (foreground)")
    .option("--host <host>", "Gateway host")
    .option("--port <port>", "Gateway port")
    .option("--tls", "Use TLS for the gateway connection", false)
    .option("--tls-fingerprint <sha256>", "Expected TLS certificate fingerprint (sha256)")
-    .option("--node-id <id>", "Override node id")
+    .option("--node-id <id>", "Override node id (clears pairing token)")
    .option("--display-name <name>", "Override node display name")
    .action(async (opts) => {
      const existing = await loadNodeHostConfig();
@@ -51,71 +50,51 @@ export function registerNodeCli(program: Command) {
      });
    });

-  const registerNodeServiceCommands = (cmd: Command) => {
-    cmd
-      .command("status")
-      .description("Show node service status")
-      .option("--json", "Output JSON", false)
-      .action(async (opts) => {
-        await runNodeDaemonStatus(opts);
-      });
+  node
+    .command("status")
+    .description("Show node host status")
+    .option("--json", "Output JSON", false)
+    .action(async (opts) => {
+      await runNodeDaemonStatus(opts);
+    });

-    cmd
-      .command("install")
-      .description("Install the node service (launchd/systemd/schtasks)")
-      .option("--host <host>", "Gateway host")
-      .option("--port <port>", "Gateway port")
-      .option("--tls", "Use TLS for the gateway connection", false)
-      .option("--tls-fingerprint <sha256>", "Expected TLS certificate fingerprint (sha256)")
-      .option("--node-id <id>", "Override node id")
-      .option("--display-name <name>", "Override node display name")
-      .option("--runtime <runtime>", "Service runtime (node|bun). Default: node")
-      .option("--force", "Reinstall/overwrite if already installed", false)
-      .option("--json", "Output JSON", false)
-      .action(async (opts) => {
-        await runNodeDaemonInstall(opts);
-      });
+  node
+    .command("install")
+    .description("Install the node host service (launchd/systemd/schtasks)")
+    .option("--host <host>", "Gateway host")
+    .option("--port <port>", "Gateway port")
+    .option("--tls", "Use TLS for the gateway connection", false)
+    .option("--tls-fingerprint <sha256>", "Expected TLS certificate fingerprint (sha256)")
+    .option("--node-id <id>", "Override node id (clears pairing token)")
+    .option("--display-name <name>", "Override node display name")
+    .option("--runtime <runtime>", "Service runtime (node|bun). Default: node")
+    .option("--force", "Reinstall/overwrite if already installed", false)
+    .option("--json", "Output JSON", false)
+    .action(async (opts) => {
+      await runNodeDaemonInstall(opts);
+    });

-    cmd
-      .command("uninstall")
-      .description("Uninstall the node service (launchd/systemd/schtasks)")
-      .option("--json", "Output JSON", false)
-      .action(async (opts) => {
-        await runNodeDaemonUninstall(opts);
-      });
+  node
+    .command("uninstall")
+    .description("Uninstall the node host service (launchd/systemd/schtasks)")
+    .option("--json", "Output JSON", false)
+    .action(async (opts) => {
+      await runNodeDaemonUninstall(opts);
+    });

-    cmd
-      .command("start")
-      .description("Start the node service (launchd/systemd/schtasks)")
-      .option("--json", "Output JSON", false)
-      .action(async (opts) => {
-        await runNodeDaemonStart(opts);
-      });
+  node
+    .command("stop")
+    .description("Stop the node host service (launchd/systemd/schtasks)")
+    .option("--json", "Output JSON", false)
+    .action(async (opts) => {
+      await runNodeDaemonStop(opts);
+    });

-    cmd
-      .command("stop")
-      .description("Stop the node service (launchd/systemd/schtasks)")
-      .option("--json", "Output JSON", false)
-      .action(async (opts) => {
-        await runNodeDaemonStop(opts);
-      });
-
-    cmd
-      .command("restart")
-      .description("Restart the node service (launchd/systemd/schtasks)")
-      .option("--json", "Output JSON", false)
-      .action(async (opts) => {
-        await runNodeDaemonRestart(opts);
-      });
-  };
-
-  const service = node
-    .command("service")
-    .description("Manage the headless node service (launchd/systemd/schtasks)");
-  registerNodeServiceCommands(service);
-
-  const daemon = node
-    .command("daemon", { hidden: true })
-    .description("Legacy alias for node service commands");
-  registerNodeServiceCommands(daemon);
+  node
+    .command("restart")
+    .description("Restart the node host service (launchd/systemd/schtasks)")
+    .option("--json", "Output JSON", false)
+    .action(async (opts) => {
+      await runNodeDaemonRestart(opts);
+    });
 }
@@ -44,6 +44,15 @@ function formatNodeVersions(node: {
  return parts.length > 0 ? parts.join(" · ") : null;
 }

+function formatPathEnv(raw?: string): string | null {
+  if (typeof raw !== "string") return null;
+  const trimmed = raw.trim();
+  if (!trimmed) return null;
+  const parts = trimmed.split(":").filter(Boolean);
+  if (parts.length <= 3) return trimmed;
+  return `${parts.slice(0, 2).join(":")}:…:${parts.slice(-1)[0]}`;
+}
+
 function parseSinceMs(raw: unknown, label: string): number | undefined {
  if (raw === undefined || raw === null) return undefined;
  const value =
@@ -126,11 +135,13 @@ export function registerNodesStatusCommands(nodes: Command) {
            const name = n.displayName?.trim() ? n.displayName.trim() : n.nodeId;
            const perms = formatPermissions(n.permissions);
            const versions = formatNodeVersions(n);
+            const pathEnv = formatPathEnv(n.pathEnv);
            const detailParts = [
              n.deviceFamily ? `device: ${n.deviceFamily}` : null,
              n.modelIdentifier ? `hw: ${n.modelIdentifier}` : null,
              perms ? `perms: ${perms}` : null,
              versions,
+              pathEnv ? `path: ${pathEnv}` : null,
            ].filter(Boolean) as string[];
            const caps = Array.isArray(n.caps)
              ? n.caps.map(String).filter(Boolean).sort().join(", ")
@@ -201,6 +212,7 @@ export function registerNodesStatusCommands(nodes: Command) {
          const family = typeof obj.deviceFamily === "string" ? obj.deviceFamily : null;
          const model = typeof obj.modelIdentifier === "string" ? obj.modelIdentifier : null;
          const ip = typeof obj.remoteIp === "string" ? obj.remoteIp : null;
+          const pathEnv = typeof obj.pathEnv === "string" ? obj.pathEnv : null;
          const versions = formatNodeVersions(
            obj as {
              platform?: string;
@@ -223,6 +235,7 @@ export function registerNodesStatusCommands(nodes: Command) {
            model ? { Field: "Model", Value: model } : null,
            perms ? { Field: "Perms", Value: perms } : null,
            versions ? { Field: "Version", Value: versions } : null,
+            pathEnv ? { Field: "PATH", Value: pathEnv } : null,
            { Field: "Status", Value: status },
            { Field: "Caps", Value: caps ? caps.join(", ") : "?" },
          ].filter(Boolean) as Array<{ Field: string; Value: string }>;
@@ -53,6 +53,7 @@ export type NodeListNode = {
  remoteIp?: string;
  deviceFamily?: string;
  modelIdentifier?: string;
+  pathEnv?: string;
  caps?: string[];
  commands?: string[];
  permissions?: Record<string, boolean>;
@@ -4,6 +4,7 @@ import { emitCliBanner } from "../banner.js";
 import { getCommandPath, getVerboseFlag, hasHelpOrVersion } from "../argv.js";
 import { ensureConfigReady } from "./config-guard.js";
 import { ensurePluginRegistryLoaded } from "../plugin-registry.js";
+import { isTruthyEnvValue } from "../../infra/env.js";
 import { setVerbose } from "../../globals.js";

 function setProcessTitleForCommand(actionCommand: Command) {
@@ -22,15 +23,21 @@ const PLUGIN_REQUIRED_COMMANDS = new Set(["message", "channels", "directory"]);
 export function registerPreActionHooks(program: Command, programVersion: string) {
  program.hook("preAction", async (_thisCommand, actionCommand) => {
    setProcessTitleForCommand(actionCommand);
-    emitCliBanner(programVersion);
    const argv = process.argv;
    if (hasHelpOrVersion(argv)) return;
+    const commandPath = getCommandPath(argv, 2);
+    const hideBanner =
+      isTruthyEnvValue(process.env.CLAWDBOT_HIDE_BANNER) ||
+      commandPath[0] === "update" ||
+      (commandPath[0] === "plugins" && commandPath[1] === "update");
+    if (!hideBanner) {
+      emitCliBanner(programVersion);
+    }
    const verbose = getVerboseFlag(argv, { includeDebug: true });
    setVerbose(verbose);
    if (!verbose) {
      process.env.NODE_NO_WARNINGS ??= "1";
    }
-    const commandPath = getCommandPath(argv, 2);
    if (commandPath[0] === "doctor") return;
    await ensureConfigReady({ runtime: defaultRuntime, commandPath });
    // Load plugins for commands that need channel access
@@ -52,14 +52,6 @@ const entries: SubCliEntry[] = [
      mod.registerGatewayCli(program);
    },
  },
-  {
-    name: "service",
-    description: "Service helpers",
-    register: async (program) => {
-      const mod = await import("../service-cli.js");
-      mod.registerServiceCli(program);
-    },
-  },
  {
    name: "logs",
    description: "Gateway logs",
@@ -1,59 +0,0 @@
-import { Command } from "commander";
-import { describe, expect, it, vi } from "vitest";
-
-const runDaemonStatus = vi.fn(async () => {});
-const runNodeDaemonStatus = vi.fn(async () => {});
-
-vi.mock("./daemon-cli/runners.js", () => ({
-  runDaemonInstall: vi.fn(async () => {}),
-  runDaemonRestart: vi.fn(async () => {}),
-  runDaemonStart: vi.fn(async () => {}),
-  runDaemonStatus: (opts: unknown) => runDaemonStatus(opts),
-  runDaemonStop: vi.fn(async () => {}),
-  runDaemonUninstall: vi.fn(async () => {}),
-}));
-
-vi.mock("./node-cli/daemon.js", () => ({
-  runNodeDaemonInstall: vi.fn(async () => {}),
-  runNodeDaemonRestart: vi.fn(async () => {}),
-  runNodeDaemonStart: vi.fn(async () => {}),
-  runNodeDaemonStatus: (opts: unknown) => runNodeDaemonStatus(opts),
-  runNodeDaemonStop: vi.fn(async () => {}),
-  runNodeDaemonUninstall: vi.fn(async () => {}),
-}));
-
-vi.mock("./deps.js", () => ({
-  createDefaultDeps: vi.fn(),
-}));
-
-describe("service CLI coverage", () => {
-  it("routes service gateway status to daemon status", async () => {
-    runDaemonStatus.mockClear();
-    runNodeDaemonStatus.mockClear();
-
-    const { registerServiceCli } = await import("./service-cli.js");
-    const program = new Command();
-    program.exitOverride();
-    registerServiceCli(program);
-
-    await program.parseAsync(["service", "gateway", "status"], { from: "user" });
-
-    expect(runDaemonStatus).toHaveBeenCalledTimes(1);
-    expect(runNodeDaemonStatus).toHaveBeenCalledTimes(0);
-  });
-
-  it("routes service node status to node daemon status", async () => {
-    runDaemonStatus.mockClear();
-    runNodeDaemonStatus.mockClear();
-
-    const { registerServiceCli } = await import("./service-cli.js");
-    const program = new Command();
-    program.exitOverride();
-    registerServiceCli(program);
-
-    await program.parseAsync(["service", "node", "status"], { from: "user" });
-
-    expect(runNodeDaemonStatus).toHaveBeenCalledTimes(1);
-    expect(runDaemonStatus).toHaveBeenCalledTimes(0);
-  });
-});
@@ -1,181 +0,0 @@
-import type { Command } from "commander";
-import { formatDocsLink } from "../terminal/links.js";
-import { theme } from "../terminal/theme.js";
-import { formatHelpExampleGroup } from "./help-format.js";
-import { createDefaultDeps } from "./deps.js";
-import {
-  runDaemonInstall,
-  runDaemonRestart,
-  runDaemonStart,
-  runDaemonStatus,
-  runDaemonStop,
-  runDaemonUninstall,
-} from "./daemon-cli/runners.js";
-import {
-  runNodeDaemonInstall,
-  runNodeDaemonRestart,
-  runNodeDaemonStart,
-  runNodeDaemonStatus,
-  runNodeDaemonStop,
-  runNodeDaemonUninstall,
-} from "./node-cli/daemon.js";
-
-export function registerServiceCli(program: Command) {
-  const gatewayExamples: Array<[string, string]> = [
-    ["clawdbot service gateway status", "Show gateway service status + probe."],
-    [
-      "clawdbot service gateway install --port 18789 --token <token>",
-      "Install the Gateway service on port 18789.",
-    ],
-    ["clawdbot service gateway restart", "Restart the Gateway service."],
-  ];
-
-  const nodeExamples: Array<[string, string]> = [
-    ["clawdbot service node status", "Show node host service status."],
-    [
-      "clawdbot service node install --host gateway.local --port 18789 --tls",
-      "Install the node host service with TLS.",
-    ],
-    ["clawdbot service node restart", "Restart the node host service."],
-  ];
-
-  const service = program
-    .command("service")
-    .description("Manage Gateway and node host services (launchd/systemd/schtasks)")
-    .addHelpText(
-      "after",
-      () =>
-        `\n${theme.heading("Examples:")}\n${formatHelpExampleGroup(
-          "Gateway:",
-          gatewayExamples,
-        )}\n\n${formatHelpExampleGroup("Node:", nodeExamples)}\n\n${theme.muted(
-          "Docs:",
-        )} ${formatDocsLink("/cli/service", "docs.clawd.bot/cli/service")}\n`,
-    );
-
-  const gateway = service.command("gateway").description("Manage the Gateway service");
-
-  gateway
-    .command("status")
-    .description("Show gateway service status + probe the Gateway")
-    .option("--url <url>", "Gateway WebSocket URL (defaults to config/remote/local)")
-    .option("--token <token>", "Gateway token (if required)")
-    .option("--password <password>", "Gateway password (password auth)")
-    .option("--timeout <ms>", "Timeout in ms", "10000")
-    .option("--no-probe", "Skip RPC probe")
-    .option("--deep", "Scan system-level services", false)
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runDaemonStatus({
-        rpc: opts,
-        probe: Boolean(opts.probe),
-        deep: Boolean(opts.deep),
-        json: Boolean(opts.json),
-      });
-    });
-
-  gateway
-    .command("install")
-    .description("Install the Gateway service (launchd/systemd/schtasks)")
-    .option("--port <port>", "Gateway port")
-    .option("--runtime <runtime>", "Service runtime (node|bun). Default: node")
-    .option("--token <token>", "Gateway token (token auth)")
-    .option("--force", "Reinstall/overwrite if already installed", false)
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runDaemonInstall(opts);
-    });
-
-  gateway
-    .command("uninstall")
-    .description("Uninstall the Gateway service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runDaemonUninstall(opts);
-    });
-
-  gateway
-    .command("start")
-    .description("Start the Gateway service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runDaemonStart(opts);
-    });
-
-  gateway
-    .command("stop")
-    .description("Stop the Gateway service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runDaemonStop(opts);
-    });
-
-  gateway
-    .command("restart")
-    .description("Restart the Gateway service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runDaemonRestart(opts);
-    });
-
-  const node = service.command("node").description("Manage the node host service");
-
-  node
-    .command("status")
-    .description("Show node host service status")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runNodeDaemonStatus(opts);
-    });
-
-  node
-    .command("install")
-    .description("Install the node host service (launchd/systemd/schtasks)")
-    .option("--host <host>", "Gateway host")
-    .option("--port <port>", "Gateway port")
-    .option("--tls", "Use TLS for the Gateway connection", false)
-    .option("--tls-fingerprint <sha256>", "Expected TLS certificate fingerprint (sha256)")
-    .option("--node-id <id>", "Override node id (clears pairing token)")
-    .option("--display-name <name>", "Override node display name")
-    .option("--runtime <runtime>", "Service runtime (node|bun). Default: node")
-    .option("--force", "Reinstall/overwrite if already installed", false)
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runNodeDaemonInstall(opts);
-    });
-
-  node
-    .command("uninstall")
-    .description("Uninstall the node host service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runNodeDaemonUninstall(opts);
-    });
-
-  node
-    .command("start")
-    .description("Start the node host service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runNodeDaemonStart(opts);
-    });
-
-  node
-    .command("stop")
-    .description("Stop the node host service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runNodeDaemonStop(opts);
-    });
-
-  node
-    .command("restart")
-    .description("Restart the node host service (launchd/systemd/schtasks)")
-    .option("--json", "Output JSON", false)
-    .action(async (opts) => {
-      await runNodeDaemonRestart(opts);
-    });
-
-  // Build default deps (parity with daemon CLI).
-  void createDefaultDeps();
-}
@@ -19,7 +19,6 @@ import {
  type UpdateStepInfo,
  type UpdateStepResult,
  type UpdateStepProgress,
-  type UpdateStepResult,
 } from "../infra/update-runner.js";
 import {
  detectGlobalInstallManagerByPresence,
@@ -4,10 +4,22 @@ import type { RuntimeEnv } from "../runtime.js";
 import type { WizardPrompter } from "../wizard/prompts.js";
 import { applyAuthChoice, resolvePreferredProviderForAuthChoice } from "./auth-choice.js";
 import { promptAuthChoiceGrouped } from "./auth-choice-prompt.js";
-import { applyPrimaryModel, promptDefaultModel } from "./model-picker.js";
+import {
+  applyModelAllowlist,
+  applyModelFallbacksFromSelection,
+  applyPrimaryModel,
+  promptDefaultModel,
+  promptModelAllowlist,
+} from "./model-picker.js";

 type GatewayAuthChoice = "off" | "token" | "password";

+const ANTHROPIC_OAUTH_MODEL_KEYS = [
+  "anthropic/claude-opus-4-5",
+  "anthropic/claude-sonnet-4-5",
+  "anthropic/claude-haiku-4-5",
+];
+
 export function buildGatewayAuthConfig(params: {
  existing?: GatewayAuthConfig;
  mode: GatewayAuthChoice;
@@ -51,19 +63,35 @@ export async function promptAuthConfig(
      setDefaultModel: true,
    });
    next = applied.config;
-    // Auth choice already set a sensible default model; skip the model picker.
-    return next;
+  } else {
+    const modelSelection = await promptDefaultModel({
+      config: next,
+      prompter,
+      allowKeep: true,
+      ignoreAllowlist: true,
+      preferredProvider: resolvePreferredProviderForAuthChoice(authChoice),
+    });
+    if (modelSelection.model) {
+      next = applyPrimaryModel(next, modelSelection.model);
+    }
  }

-  const modelSelection = await promptDefaultModel({
+  const anthropicOAuth =
+    authChoice === "claude-cli" ||
+    authChoice === "setup-token" ||
+    authChoice === "token" ||
+    authChoice === "oauth";
+
+  const allowlistSelection = await promptModelAllowlist({
    config: next,
    prompter,
-    allowKeep: true,
-    ignoreAllowlist: true,
-    preferredProvider: resolvePreferredProviderForAuthChoice(authChoice),
+    allowedKeys: anthropicOAuth ? ANTHROPIC_OAUTH_MODEL_KEYS : undefined,
+    initialSelections: anthropicOAuth ? ["anthropic/claude-opus-4-5"] : undefined,
+    message: anthropicOAuth ? "Anthropic OAuth models" : undefined,
  });
-  if (modelSelection.model) {
-    next = applyPrimaryModel(next, modelSelection.model);
+  if (allowlistSelection.models) {
+    next = applyModelAllowlist(next, allowlistSelection.models);
+    next = applyModelFallbacksFromSelection(next, allowlistSelection.models);
  }

  return next;
@@ -2,7 +2,12 @@ import { describe, expect, it, vi } from "vitest";

 import type { ClawdbotConfig } from "../config/config.js";
 import { makePrompter } from "./onboarding/__tests__/test-utils.js";
-import { promptDefaultModel } from "./model-picker.js";
+import {
+  applyModelAllowlist,
+  applyModelFallbacksFromSelection,
+  promptDefaultModel,
+  promptModelAllowlist,
+} from "./model-picker.js";

 const loadModelCatalog = vi.hoisted(() => vi.fn());
 vi.mock("../agents/model-catalog.js", () => ({
@@ -65,3 +70,145 @@ describe("promptDefaultModel", () => {
    );
  });
 });
+
+describe("promptModelAllowlist", () => {
+  it("filters internal router models from the selection list", async () => {
+    loadModelCatalog.mockResolvedValue([
+      {
+        provider: "openrouter",
+        id: "auto",
+        name: "OpenRouter Auto",
+      },
+      {
+        provider: "openrouter",
+        id: "meta-llama/llama-3.3-70b:free",
+        name: "Llama 3.3 70B",
+      },
+    ]);
+
+    const multiselect = vi.fn(async (params) =>
+      params.options.map((option: { value: string }) => option.value),
+    );
+    const prompter = makePrompter({ multiselect });
+    const config = { agents: { defaults: {} } } as ClawdbotConfig;
+
+    await promptModelAllowlist({ config, prompter });
+
+    const options = multiselect.mock.calls[0]?.[0]?.options ?? [];
+    expect(options.some((opt: { value: string }) => opt.value === "openrouter/auto")).toBe(false);
+    expect(
+      options.some(
+        (opt: { value: string }) => opt.value === "openrouter/meta-llama/llama-3.3-70b:free",
+      ),
+    ).toBe(true);
+  });
+
+  it("filters to allowed keys when provided", async () => {
+    loadModelCatalog.mockResolvedValue([
+      {
+        provider: "anthropic",
+        id: "claude-opus-4-5",
+        name: "Claude Opus 4.5",
+      },
+      {
+        provider: "anthropic",
+        id: "claude-sonnet-4-5",
+        name: "Claude Sonnet 4.5",
+      },
+      {
+        provider: "openai",
+        id: "gpt-5.2",
+        name: "GPT-5.2",
+      },
+    ]);
+
+    const multiselect = vi.fn(async (params) =>
+      params.options.map((option: { value: string }) => option.value),
+    );
+    const prompter = makePrompter({ multiselect });
+    const config = { agents: { defaults: {} } } as ClawdbotConfig;
+
+    await promptModelAllowlist({
+      config,
+      prompter,
+      allowedKeys: ["anthropic/claude-opus-4-5"],
+    });
+
+    const options = multiselect.mock.calls[0]?.[0]?.options ?? [];
+    expect(options.map((opt: { value: string }) => opt.value)).toEqual([
+      "anthropic/claude-opus-4-5",
+    ]);
+  });
+});
+
+describe("applyModelAllowlist", () => {
+  it("preserves existing entries for selected models", () => {
+    const config = {
+      agents: {
+        defaults: {
+          models: {
+            "openai/gpt-5.2": { alias: "gpt" },
+            "anthropic/claude-opus-4-5": { alias: "opus" },
+          },
+        },
+      },
+    } as ClawdbotConfig;
+
+    const next = applyModelAllowlist(config, ["openai/gpt-5.2"]);
+    expect(next.agents?.defaults?.models).toEqual({
+      "openai/gpt-5.2": { alias: "gpt" },
+    });
+  });
+
+  it("clears the allowlist when no models remain", () => {
+    const config = {
+      agents: {
+        defaults: {
+          models: {
+            "openai/gpt-5.2": { alias: "gpt" },
+          },
+        },
+      },
+    } as ClawdbotConfig;
+
+    const next = applyModelAllowlist(config, []);
+    expect(next.agents?.defaults?.models).toBeUndefined();
+  });
+});
+
+describe("applyModelFallbacksFromSelection", () => {
+  it("sets fallbacks from selection when the primary is included", () => {
+    const config = {
+      agents: {
+        defaults: {
+          model: { primary: "anthropic/claude-opus-4-5" },
+        },
+      },
+    } as ClawdbotConfig;
+
+    const next = applyModelFallbacksFromSelection(config, [
+      "anthropic/claude-opus-4-5",
+      "anthropic/claude-sonnet-4-5",
+    ]);
+    expect(next.agents?.defaults?.model).toEqual({
+      primary: "anthropic/claude-opus-4-5",
+      fallbacks: ["anthropic/claude-sonnet-4-5"],
+    });
+  });
+
+  it("keeps existing fallbacks when the primary is not selected", () => {
+    const config = {
+      agents: {
+        defaults: {
+          model: { primary: "anthropic/claude-opus-4-5", fallbacks: ["openai/gpt-5.2"] },
+        },
+      },
+    } as ClawdbotConfig;
+
+    const next = applyModelFallbacksFromSelection(config, ["openai/gpt-5.2"]);
+    expect(next.agents?.defaults?.model).toEqual({
+      primary: "anthropic/claude-opus-4-5",
+      fallbacks: ["openai/gpt-5.2"],
+    });
+  });
+});
@@ -34,6 +34,7 @@ type PromptDefaultModelParams = {
 };

 type PromptDefaultModelResult = { model?: string };
+type PromptModelAllowlistResult = { models?: string[] };

 function hasAuthForProvider(
  provider: string,
@@ -52,6 +53,25 @@ function resolveConfiguredModelRaw(cfg: ClawdbotConfig): string {
  return raw?.primary?.trim() ?? "";
 }

+function resolveConfiguredModelKeys(cfg: ClawdbotConfig): string[] {
+  const models = cfg.agents?.defaults?.models ?? {};
+  return Object.keys(models)
+    .map((key) => String(key ?? "").trim())
+    .filter((key) => key.length > 0);
+}
+
+function normalizeModelKeys(values: string[]): string[] {
+  const seen = new Set<string>();
+  const next: string[] = [];
+  for (const raw of values) {
+    const value = String(raw ?? "").trim();
+    if (!value || seen.has(value)) continue;
+    seen.add(value);
+    next.push(value);
+  }
+  return next;
+}
+
 async function promptManualModel(params: {
  prompter: WizardPrompter;
  allowBlank: boolean;
@@ -245,6 +265,128 @@ export async function promptDefaultModel(
  return { model: String(selection) };
 }

+export async function promptModelAllowlist(params: {
+  config: ClawdbotConfig;
+  prompter: WizardPrompter;
+  message?: string;
+  agentDir?: string;
+  allowedKeys?: string[];
+  initialSelections?: string[];
+}): Promise<PromptModelAllowlistResult> {
+  const cfg = params.config;
+  const existingKeys = resolveConfiguredModelKeys(cfg);
+  const allowedKeys = normalizeModelKeys(params.allowedKeys ?? []);
+  const allowedKeySet = allowedKeys.length > 0 ? new Set(allowedKeys) : null;
+  const resolved = resolveConfiguredModelRef({
+    cfg,
+    defaultProvider: DEFAULT_PROVIDER,
+    defaultModel: DEFAULT_MODEL,
+  });
+  const resolvedKey = modelKey(resolved.provider, resolved.model);
+  const initialSeeds = normalizeModelKeys([
+    ...existingKeys,
+    resolvedKey,
+    ...(params.initialSelections ?? []),
+  ]);
+  const initialKeys = allowedKeySet
+    ? initialSeeds.filter((key) => allowedKeySet.has(key))
+    : initialSeeds;
+
+  const catalog = await loadModelCatalog({ config: cfg, useCache: false });
+  if (catalog.length === 0 && allowedKeys.length === 0) {
+    const raw = await params.prompter.text({
+      message:
+        params.message ??
+        "Allowlist models (comma-separated provider/model; blank to keep current)",
+      initialValue: existingKeys.join(", "),
+      placeholder: "openai-codex/gpt-5.2, anthropic/claude-opus-4-5",
+    });
+    const parsed = String(raw ?? "")
+      .split(",")
+      .map((value) => value.trim())
+      .filter((value) => value.length > 0);
+    if (parsed.length === 0) return {};
+    return { models: normalizeModelKeys(parsed) };
+  }
+
+  const aliasIndex = buildModelAliasIndex({
+    cfg,
+    defaultProvider: DEFAULT_PROVIDER,
+  });
+  const authStore = ensureAuthProfileStore(params.agentDir, {
+    allowKeychainPrompt: false,
+  });
+  const authCache = new Map<string, boolean>();
+  const hasAuth = (provider: string) => {
+    const cached = authCache.get(provider);
+    if (cached !== undefined) return cached;
+    const value = hasAuthForProvider(provider, cfg, authStore);
+    authCache.set(provider, value);
+    return value;
+  };
+
+  const options: WizardSelectOption<string>[] = [];
+  const seen = new Set<string>();
+  const addModelOption = (entry: {
+    provider: string;
+    id: string;
+    name?: string;
+    contextWindow?: number;
+    reasoning?: boolean;
+  }) => {
+    const key = modelKey(entry.provider, entry.id);
+    if (seen.has(key)) return;
+    if (HIDDEN_ROUTER_MODELS.has(key)) return;
+    const hints: string[] = [];
+    if (entry.name && entry.name !== entry.id) hints.push(entry.name);
+    if (entry.contextWindow) hints.push(`ctx ${formatTokenK(entry.contextWindow)}`);
+    if (entry.reasoning) hints.push("reasoning");
+    const aliases = aliasIndex.byKey.get(key);
+    if (aliases?.length) hints.push(`alias: ${aliases.join(", ")}`);
+    if (!hasAuth(entry.provider)) hints.push("auth missing");
+    options.push({
+      value: key,
+      label: key,
+      hint: hints.length > 0 ? hints.join(" · ") : undefined,
+    });
+    seen.add(key);
+  };
+
+  const filteredCatalog = allowedKeySet
+    ? catalog.filter((entry) => allowedKeySet.has(modelKey(entry.provider, entry.id)))
+    : catalog;
+
+  for (const entry of filteredCatalog) addModelOption(entry);
+
+  const supplementalKeys = allowedKeySet ? allowedKeys : existingKeys;
+  for (const key of supplementalKeys) {
+    if (seen.has(key)) continue;
+    options.push({
+      value: key,
+      label: key,
+      hint: allowedKeySet ? "allowed (not in catalog)" : "configured (not in catalog)",
+    });
+    seen.add(key);
+  }
+
+  if (options.length === 0) return {};
+
+  const selection = await params.prompter.multiselect({
+    message: params.message ?? "Models in /model picker (multi-select)",
+    options,
+    initialValues: initialKeys.length > 0 ? initialKeys : undefined,
+  });
+  const selected = normalizeModelKeys(selection.map((value) => String(value)));
+  if (selected.length > 0) return { models: selected };
+  if (existingKeys.length === 0) return { models: [] };
+  const confirmClear = await params.prompter.confirm({
+    message: "Clear the model allowlist? (shows all models)",
+    initialValue: false,
+  });
+  if (!confirmClear) return {};
+  return { models: [] };
+}
+
 export function applyPrimaryModel(cfg: ClawdbotConfig, model: string): ClawdbotConfig {
  const defaults = cfg.agents?.defaults;
  const existingModel = defaults?.model;
@@ -271,3 +413,77 @@ export function applyPrimaryModel(cfg: ClawdbotConfig, model: string): ClawdbotC
    },
  };
 }
+
+export function applyModelAllowlist(cfg: ClawdbotConfig, models: string[]): ClawdbotConfig {
+  const defaults = cfg.agents?.defaults;
+  const normalized = normalizeModelKeys(models);
+  if (normalized.length === 0) {
+    if (!defaults?.models) return cfg;
+    const { models: _ignored, ...restDefaults } = defaults;
+    return {
+      ...cfg,
+      agents: {
+        ...cfg.agents,
+        defaults: restDefaults,
+      },
+    };
+  }
+
+  const existingModels = defaults?.models ?? {};
+  const nextModels: Record<string, { alias?: string }> = {};
+  for (const key of normalized) {
+    nextModels[key] = existingModels[key] ?? {};
+  }
+
+  return {
+    ...cfg,
+    agents: {
+      ...cfg.agents,
+      defaults: {
+        ...defaults,
+        models: nextModels,
+      },
+    },
+  };
+}
+
+export function applyModelFallbacksFromSelection(
+  cfg: ClawdbotConfig,
+  selection: string[],
+): ClawdbotConfig {
+  const normalized = normalizeModelKeys(selection);
+  if (normalized.length <= 1) return cfg;
+
+  const resolved = resolveConfiguredModelRef({
+    cfg,
+    defaultProvider: DEFAULT_PROVIDER,
+    defaultModel: DEFAULT_MODEL,
+  });
+  const resolvedKey = modelKey(resolved.provider, resolved.model);
+  if (!normalized.includes(resolvedKey)) return cfg;
+
+  const defaults = cfg.agents?.defaults;
+  const existingModel = defaults?.model;
+  const existingPrimary =
+    typeof existingModel === "string"
+      ? existingModel
+      : existingModel && typeof existingModel === "object"
+        ? existingModel.primary
+        : undefined;
+
+  const fallbacks = normalized.filter((key) => key !== resolvedKey);
+  return {
+    ...cfg,
+    agents: {
+      ...cfg.agents,
+      defaults: {
+        ...defaults,
+        model: {
+          ...(typeof existingModel === "object" ? existingModel : undefined),
+          primary: existingPrimary ?? resolvedKey,
+          fallbacks,
+        },
+      },
+    },
+  };
+}
@@ -114,6 +114,11 @@ const FIELD_LABELS: Record<string, string> = {
  "diagnostics.otel.logs": "OpenTelemetry Logs Enabled",
  "diagnostics.otel.sampleRate": "OpenTelemetry Trace Sample Rate",
  "diagnostics.otel.flushIntervalMs": "OpenTelemetry Flush Interval (ms)",
+  "diagnostics.cacheTrace.enabled": "Cache Trace Enabled",
+  "diagnostics.cacheTrace.filePath": "Cache Trace File Path",
+  "diagnostics.cacheTrace.includeMessages": "Cache Trace Include Messages",
+  "diagnostics.cacheTrace.includePrompt": "Cache Trace Include Prompt",
+  "diagnostics.cacheTrace.includeSystem": "Cache Trace Include System",
  "gateway.remote.url": "Remote Gateway URL",
  "gateway.remote.sshTarget": "Remote Gateway SSH Target",
  "gateway.remote.sshIdentity": "Remote Gateway SSH Identity",
@@ -215,6 +220,8 @@ const FIELD_LABELS: Record<string, string> = {
  "agents.defaults.memorySearch.sync.onSearch": "Index on Search (Lazy)",
  "agents.defaults.memorySearch.sync.watch": "Watch Memory Files",
  "agents.defaults.memorySearch.sync.watchDebounceMs": "Memory Watch Debounce (ms)",
+  "agents.defaults.memorySearch.sync.sessions.deltaBytes": "Session Delta Bytes",
+  "agents.defaults.memorySearch.sync.sessions.deltaMessages": "Session Delta Messages",
  "agents.defaults.memorySearch.query.maxResults": "Memory Search Max Results",
  "agents.defaults.memorySearch.query.minScore": "Memory Search Min Score",
  "agents.defaults.memorySearch.query.hybrid.enabled": "Memory Search Hybrid",
@@ -345,6 +352,14 @@ const FIELD_HELP: Record<string, string> = {
    "Extra node.invoke commands to allow beyond the gateway defaults (array of command strings).",
  "gateway.nodes.denyCommands":
    "Commands to block even if present in node claims or default allowlist.",
+  "diagnostics.cacheTrace.enabled":
+    "Log cache trace snapshots for embedded agent runs (default: false).",
+  "diagnostics.cacheTrace.filePath":
+    "JSONL output path for cache trace logs (default: $CLAWDBOT_STATE_DIR/logs/cache-trace.jsonl).",
+  "diagnostics.cacheTrace.includeMessages":
+    "Include full message payloads in trace output (default: true).",
+  "diagnostics.cacheTrace.includePrompt": "Include prompt text in trace output (default: true).",
+  "diagnostics.cacheTrace.includeSystem": "Include system prompt in trace output (default: true).",
  "tools.exec.applyPatch.enabled":
    "Experimental. Enables apply_patch for OpenAI models when allowed by tool policy.",
  "tools.exec.applyPatch.allowModels":
@@ -461,8 +476,12 @@ const FIELD_HELP: Record<string, string> = {
  "agents.defaults.memorySearch.cache.maxEntries":
    "Optional cap on cached embeddings (best-effort).",
  "agents.defaults.memorySearch.sync.onSearch":
-    "Lazy sync: reindex on first search after a change.",
+    "Lazy sync: schedule a reindex on search after changes.",
  "agents.defaults.memorySearch.sync.watch": "Watch memory files for changes (chokidar).",
+  "agents.defaults.memorySearch.sync.sessions.deltaBytes":
+    "Minimum appended bytes before session transcripts trigger reindex (default: 100000).",
+  "agents.defaults.memorySearch.sync.sessions.deltaMessages":
+    "Minimum appended JSONL lines before session transcripts trigger reindex (default: 50).",
  "plugins.enabled": "Enable plugin/extension loading (default: true).",
  "plugins.allow": "Optional allowlist of plugin ids; when set, only listed plugins load.",
  "plugins.deny": "Optional denylist of plugin ids; deny wins over allowlist.",
@@ -117,9 +117,18 @@ export type DiagnosticsOtelConfig = {
  flushIntervalMs?: number;
 };

+export type DiagnosticsCacheTraceConfig = {
+  enabled?: boolean;
+  filePath?: string;
+  includeMessages?: boolean;
+  includePrompt?: boolean;
+  includeSystem?: boolean;
+};
+
 export type DiagnosticsConfig = {
  enabled?: boolean;
  otel?: DiagnosticsOtelConfig;
+  cacheTrace?: DiagnosticsCacheTraceConfig;
 };

 export type WebReconnectConfig = {
@@ -244,6 +244,12 @@ export type MemorySearchConfig = {
    watch?: boolean;
    watchDebounceMs?: number;
    intervalMinutes?: number;
+    sessions?: {
+      /** Minimum appended bytes before session transcripts are reindexed. */
+      deltaBytes?: number;
+      /** Minimum appended JSONL lines before session transcripts are reindexed. */
+      deltaMessages?: number;
+    };
  };
  /** Query behavior. */
  query?: {
@@ -305,6 +305,13 @@ export const MemorySearchSchema = z
        watch: z.boolean().optional(),
        watchDebounceMs: z.number().int().nonnegative().optional(),
        intervalMinutes: z.number().int().nonnegative().optional(),
+        sessions: z
+          .object({
+            deltaBytes: z.number().int().nonnegative().optional(),
+            deltaMessages: z.number().int().nonnegative().optional(),
+          })
+          .strict()
+          .optional(),
      })
      .strict()
      .optional(),
@@ -63,6 +63,16 @@ export const ClawdbotSchema = z
          })
          .strict()
          .optional(),
+        cacheTrace: z
+          .object({
+            enabled: z.boolean().optional(),
+            filePath: z.string().optional(),
+            includeMessages: z.boolean().optional(),
+            includePrompt: z.boolean().optional(),
+            includeSystem: z.boolean().optional(),
+          })
+          .strict()
+          .optional(),
      })
      .strict()
      .optional(),
@@ -53,6 +53,7 @@ export type GatewayClientOptions = {
  caps?: string[];
  commands?: string[];
  permissions?: Record<string, boolean>;
+  pathEnv?: string;
  deviceIdentity?: DeviceIdentity;
  minProtocol?: number;
  maxProtocol?: number;
@@ -232,6 +233,7 @@ export class GatewayClient {
        this.opts.permissions && typeof this.opts.permissions === "object"
          ? this.opts.permissions
          : undefined,
+      pathEnv: this.opts.pathEnv,
      auth,
      role,
      scopes,
Author	SHA1	Message	Date
Peter Steinberger	56fab1df69	fix: harden bluebubbles short ids and fetch wrapper (#1369 ) (thanks @tyler6204)	2026-01-21 17:05:36 +00:00
Tyler Yust	bf529f56b4	feat: enhance message context with full ID support for replies and caching - Updated message processing to include full message IDs alongside short IDs for better context resolution. - Improved reply handling by caching inbound messages, allowing for accurate sender and body resolution without exposing dropped content. - Adjusted tests to validate the new full ID properties and their integration into the message handling workflow.	2026-01-21 16:50:29 +00:00
Tyler Yust	1ada67ffb3	feat: enhance message handling with short ID resolution and reply context improvements - Implemented resolution of short message IDs to full UUIDs in both text and media sending functions. - Updated reply context formatting to optimize token usage by including only necessary information. - Introduced truncation for long reply bodies to further reduce token consumption. - Adjusted tests to reflect changes in reply context handling and message ID resolution.	2026-01-21 16:50:28 +00:00
Tyler Yust	c607dfc1d6	feat: implement short ID mapping for BlueBubbles messages and enhance reply context caching - Added functionality to resolve short message IDs to full UUIDs and vice versa, optimizing token usage. - Introduced a reply cache to store message context for replies when metadata is omitted in webhook payloads. - Updated message handling to utilize short IDs for outbound messages and replies, improving efficiency. - Enhanced error messages to clarify required parameters for actions like react, edit, and unsend. - Added tests to ensure correct behavior of new features and maintain existing functionality.	2026-01-21 16:50:28 +00:00
Peter Steinberger	39e24c9937	docs: update node CLI references	2026-01-21 16:48:42 +00:00
Peter Steinberger	fa1bc589e4	feat: flatten node CLI commands	2026-01-21 16:48:42 +00:00
Peter Steinberger	0e003cb7f1	fix: normalize abort signals for telegram fetch	2026-01-21 16:46:58 +00:00
Peter Steinberger	fb164b321e	fix: model picker allowlist fallbacks	2026-01-21 11:22:33 +00:00
Peter Steinberger	884211a924	feat: render approvals tables on write	2026-01-21 11:10:03 +00:00
Peter Steinberger	9bd6b3fd54	feat: show node PATH and bootstrap node host env	2026-01-21 11:06:56 +00:00
Peter Steinberger	dc06b225cd	fix: narrow configure model allowlist for Anthropic OAuth	2026-01-21 11:00:28 +00:00
Peter Steinberger	cdb35c3aae	test: stabilize exec approvals homedir	2026-01-21 10:49:12 +00:00
Peter Steinberger	4e4f5558fc	fix: limit /model list output	2026-01-21 10:47:37 +00:00
Peter Steinberger	8479dc97da	fix: make session memory indexing async	2026-01-21 10:39:00 +00:00
Peter Steinberger	86ddd3c69c	Merge pull request #1370 from parubets/fix-debug-ttl-cache WIP: cache trace mvp for Anthropic	2026-01-21 10:35:31 +00:00
Peter Steinberger	49d53ff0bb	fix: honor wildcard exec approvals on macOS	2026-01-21 10:27:19 +00:00
Peter Steinberger	97e8f9d619	fix: add diagnostics cache trace config (#1370 ) (thanks @parubets)	2026-01-21 10:23:30 +00:00
Andrii	5392fa0dfa	cache trace mvp Added a standalone cache tracing module and wired it into the embedded runner so you can capture message flow and the exact context sent to Anthropic in a separate JSONL file. What changed - New tracing module: src/agents/cache-trace.ts (self‑contained, env‑gated, writes JSONL, computes per‑message digests). - Hook points in src/agents/pi-embedded-runner/run/attempt.ts: logs stage snapshots (loaded/sanitized/limited/prompt/stream/after) and wraps the stream fn to record the real context.messages at send time. How to enable - CLAWDBOT_CACHE_TRACE=1 enables tracing. - CLAWDBOT_CACHE_TRACE_FILE=~/.clawdbot/logs/cache-trace.jsonl overrides output (default is $CLAWDBOT_STATE_DIR/logs/cache-trace.jsonl). - Optional filters: - CLAWDBOT_CACHE_TRACE_MESSAGES=0 to omit full messages (still logs digests). - CLAWDBOT_CACHE_TRACE_PROMPT=0 to omit prompt text. - CLAWDBOT_CACHE_TRACE_SYSTEM=0 to omit system prompt. What you’ll see - One JSON object per line with stage, messagesDigest, per‑message messageFingerprints, and the actual messages if enabled. - The most important line is stage: "stream:context" — that is the exact payload pi‑mono is sending. If this diverges from earlier stages, you’ve found the mutation point.	2026-01-21 10:23:30 +00:00
Peter Steinberger	63d017c3af	fix: add node tool failure context	2026-01-21 09:55:10 +00:00
Peter Steinberger	40646c73af	feat: improve exec approvals defaults and wildcard	2026-01-21 09:55:10 +00:00
Peter Steinberger	43ea7665ef	chore: bump bluebubbles to 2026.1.21-1	2026-01-21 09:43:19 +00:00
Peter Steinberger	ba131b0164	Update README to remove contributor acknowledgments Removed special thanks and core contributors section.	2026-01-21 09:21:50 +00:00
Peter Steinberger	0693c7804f	test: skip plugin tools in sessions tools test	2026-01-21 09:17:27 +00:00
Peter Steinberger	6c69ea2c91	refactor: centralize sandbox runtime label	2026-01-21 09:07:21 +00:00
Peter Steinberger	1e10dc1d3b	fix: use plugin-sdk exports in bluebubbles	2026-01-21 09:03:09 +00:00
Peter Steinberger	c22a37976d	fix: report sandboxed runtime in status	2026-01-21 08:59:32 +00:00
Peter Steinberger	89c5035aa2	docs: restore 2026.1.20 release notes	2026-01-21 08:43:05 +00:00
Peter Steinberger	cb7791c8a4	chore: release 2026.1.20-2	2026-01-21 08:30:33 +00:00
Peter Steinberger	9a14267dfa	chore: update appcast	2026-01-21 08:25:20 +00:00
Peter Steinberger	010d305401	chore: tidy package files list	2026-01-21 08:25:01 +00:00
Peter Steinberger	3210c91f6b	chore: release 2026.1.20	2026-01-21 08:23:49 +00:00
Peter Steinberger	e3cea55d72	docs: add npm files check to release checklist	2026-01-21 08:10:53 +00:00
Peter Steinberger	687a902f3e	fix: align chat composer	2026-01-21 07:48:00 +00:00
Peter Steinberger	fe860de148	fix: quiet update banner and skip duplicate plugin CLI	2026-01-21 07:37:22 +00:00
Peter Steinberger	bc8a59faa4	chore: release 2026.1.20-1	2026-01-21 07:37:22 +00:00
Peter Steinberger	91bcdad503	fix: guard anthropic refusal trigger	2026-01-21 07:28:49 +00:00
Peter Steinberger	ab97c6880b	Merge pull request #1360 from SocialNerd42069/fix/duplicate-assistant-texts fix: prevent duplicate assistant texts from whitespace differences	2026-01-21 06:31:01 +00:00